3892 PDF

The Algebraic and Geometric Theory of
Quadratic Forms
Richard Elman, Nikita Karpenko,

and Alexander Merkurjev
Contents
Introduction vii
Part . Classical theory of symmetric bilinear forms and quadratic forms 1

Chapter I. Bilinear Forms 3
1. Foundations 3
2. The Witt and Witt-Grothendieck rings of symmetric bilinear forms 11
3. Chain equivalence 14
4. Structure of the Witt ring 15
5. The Stiefel-Whitney map 21
6. Bilinear Pfister forms 25
Chapter II. Quadratic Forms 33
7. Foundations 33
8. Witt’s Theorems 41
9. Quadratic Pfister forms I 46
10. Totally singular forms 51
11. The Clifford algebra 52
12. Binary quadratic forms and quadratic algebras 55
13. The discriminant 57
14. The Clifford invariant 59
15. Chain p-equivalence of quadratic Pfister forms 59
16. Cohomological invariants 63
Chapter III. Forms over Rational Function Fields 67
17. The Cassels-Pfister Theorem 67
18. Values of forms 71
19. Forms over a discrete valuation ring 75
20. Similarities of forms ¡ ¢ 78
21. An exact sequence for W F (t) 85
Chapter IV. Function Fields of Quadrics 91
22. Quadrics 91
23. Quadratic Pfister forms II 96
24. Linkage of quadratic forms 99
25. The submodule Jn (F ) 102
26. The Separation Theorem 105
27. A further characterization of quadratic Pfister forms 108
iii
iv CONTENTS
28. Excellent quadratic forms 110

29. Excellent field extensions 112
30. Central simple algebras over function fields of quadratic forms 115
Chapter V. Bilinear and Quadratic Forms and Algebraic Extensions 121
31. Structure of the Witt ring 121
32. Addendum on torsion 132
33. The total signature 134
34. Bilinear and quadratic forms under quadratic extensions 138
35. Torsion in I n (F ) and torsion Pfister forms 148
Chapter VI. u-invariants 163
36. The ū-invariant 163
37. The u-invariant for formally real fields 167
38. Construction of fields with even u-invariant 172
39. Addendum: Linked fields and the Hasse number 174
Chapter VII. Applications of the Milnor Conjecture 179
40. Exact sequences for quadratic extensions 179
41. Annihilators of Pfister forms 183
42. Presentation of I n (F ) 186
43. Going down and torsion-freeness 190
Chapter VIII. On the norm residue homomorphism of degree two 195
44. The main theorem 195
45. Geometry of conic curves 196
46. Key exact sequence 200
47. Hilbert theorem 90 for K2 212
48. Proof of the main theorem 215
Part . Algebraic cycles 219

Chapter IX. Homology and cohomology 221
49. The complex C∗ (X) 221
50. External products 236
51. Deformation homomorphisms 240
52. K-homology groups 244
53. Euler classes and Projective Bundle Theorem 248
54. Chern classes 252
55. Gysin and pull-back homomorphisms 256
56. K-cohomology ring of smooth schemes 262
Chapter X. Chow groups 267
57. Definition of Chow groups 267
58. Segre and Chern classes 274
Chapter XI. Steenrod operations 283
59. Definition of the Steenrod operations 284
CONTENTS v
60. Properties of the Steenrod operations 287

61. Steenrod operations on smooth schemes 289
Chapter XII. Category of Chow motives 297
62. Correspondences 297
63. Categories of correspondences 301
64. Category of Chow motives 304
65. Duality 305
66. Motives of cellular schemes 306
67. Nilpotence Theorem 308
Part . Quadratic forms and algebraic cycles 311

Chapter XIII. Cycles on powers of quadrics 313
68. Split quadrics 313
69. Isomorphisms of quadrics 315
70. Isotropic quadrics 316
71. The Chow group of dimension 0 cycles on quadrics 317
72. The reduced Chow group 319
73. Cycles on X 2 322
Chapter XIV. The Izhboldin dimension 333
74. The first Witt index of subforms 333
75. Correspondences 334
76. The main theorem 337
77. Addendum: The Pythagoras number 340
Chapter XV. Application of Steenrod operations 343
78. Computation of Steenrod operations 343
79. Values of the first Witt index 344
80. Rost correspondences 347
81. On the 2-adic order of higher Witt indices, I 350
82. Holes in I n 355
83. On the 2-adic order of higher Witt indices, II 358
84. Minimal height 359
Chapter XVI. The variety of maximal totally isotropic subspaces 365
85. The variety Gr(ϕ) 365
86. The Chow ring of Gr(ϕ) in the split case 366
87. The Chow ring of Gr(ϕ) in the general case 371
88. The invariant J(ϕ) ¡ ¢ 374
89. Steenrod operations on Ch Gr(ϕ) 376
90. Canonical dimension 377
Chapter XVII. Motives of quadrics 381
91. Comparison of some discrete invariants of quadratic forms 381
92. The Nilpotence Theorem for quadrics 383
vi CONTENTS
93. Criterion of isomorphism 385

94. Indecomposable summands 388
Appendices 391
95. Formally real fields 392
96. The space of orderings 393
97. Cn -fields 394
98. Algebras 396
99. Galois cohomology 402
100. Milnor K-theory of fields 407
101. The cohomology groups H n,i (F, Z/mZ) 412
102. Length and Herbrand index 417
103. Places 418
104. Cones and vector bundles 420
105. Group actions on algebraic schemes 429
Bibliography 433
Index 440
Notation 447
INTRODUCTION vii
Introduction
The algebraic theory of quadratic forms, i.e., the study of quadratic forms over ar-
bitrary fields, really began with the pioneering work of Witt. In his paper [146], Witt
considered the totality of non-degenerate symmetric bilinear forms over an arbitrary field
F of characteristic different from two. Under this assumption, the theory of symmetric
bilinear forms and the theory of quadratic forms are essentially the same.
His work allowed him to form a ring W (F ), now called the Witt ring, arising from
the isometry classes of such forms. This work set the stage for further study. From the
viewpoint of ring theory, Witt gave a presentation of this ring as a quotient of the integral
group ring where the group consists of the non-zero square classes of the field F . Three
methods of study arise: ring theoretic, field theoretic, i.e., the relationship of W (F ) and
W (K) where K is a field extension of F , and algebraic geometric. In this book, we will
develop all three methods. Historically, the powerful approach using algebraic geometry
has been the last to be developed. This volume attempts to show its usefulness.
The theory of quadratic forms lay dormant until work of Cassels and then of Pfister
in the 1960’s still under the assumption of the field being of characteristic different from
two. Pfister employed the first two methods, ring theoretic and field theoretic, as well as a
nascent algebraic geometric approach. In his Habilitationsschrift [113] Pfister determined
many properties of the Witt ring. His study bifurcated into two cases: formally real fields,
i.e., fields in which −1 is not a sum of squares and non-formally real fields. In particular,
the Krull dimension of the Witt ring is one in the formally real case and zero otherwise.
This makes the study of the interaction of bilinear forms and orderings an imperative
hence the importance of looking at real closures of the base field resulting in extensions of
Sylvester’s work and Artin-Schreier theory. Pfister determined the radical, zero-divisors,
and spectrum of the Witt ring. Even earlier, in [111], he discovered remarkable forms,
now called Pfister forms. These are forms that are tensor products of binary forms that
represent one. Pfister showed that scalar multiples of these were precisely the forms that
become hyperbolic over their function field. In addition, the non-zero value set of a Pfister
form is a group and in fact the group of similarity factors of the form. As an example,
this applies to the quadratic form that is a sum of 2n squares. Pfister also used it to
show that in a non formally real field, the least number s(F ) so that −1 is a sum of s(F )
squares is always a power of 2 (cf. [112]). Interest and problems about other arithmetic
field invariants have also played a role in the development of the theory.
The non-degenerate even dimensional symmetric bilinear forms determine an ¡ ideal
¢n
n
I(F ) in the Witt ring of F , called the fundamental ideal. Its powers I (F ) := I(F )
give an important filtration of W (F ), each generated by appropriate Pfister forms. The
problem then arises: What ring theoretic properties respect this filtration? From W (F )
one also forms the graded ring GW (F ) associated to I(F ) and asks the same question.
Using Matsumoto’s presentation of K2 (F ) of a field (cf. [99]), Milnor gave an ad hoc
definition of a graded ring K∗ (F ) := ⊕n≥0 Kn (F ) of a field in [108]. From the viewpoint
of Galois cohomology, this was of great interest as there is a natural map, called the
norm residue map from Kn (F ) to the Galois cohomology group H n (ΓF , µ⊗n m ) where ΓF is
the absolute Galois group of F and m is relatively prime to the characteristic of F . For
viii CONTENTS
the case m = 2, Milnor conjectured this map to be an epimorphism with kernel 2Kn (F )
for all n. Voevodsky proved this conjecture in [142]. Milnor also related his algebraic
K-ring of a field to quadratic form theory, by asking if GW (F ) and K∗ (F )/2K∗ (F ) were
isomorphic. This was solved in the affirmative by Orlov, Vishik, and Voevodsky in [110].
Assuming these results, one can answer some of the questions that have arisen about the
filtration of W (F ) induced by the fundamental ideal.
In this book, we do not restrict ourselves to fields of characteristic different from two.
Historically the cases of fields of characteristic different from two and two have been
studied separately. Usually the case of characteristic different from two investigated first.
In this book, we shall give characteristic free proofs whenever possible. This means that
the study of symmetric bilinear forms and the study of quadratic forms must be done
separately, then interrelated. Not only do we present the classical theory characteristic
free but include many results not proven in any text as well as some previously unpublished
results to bring the classical theory up to date.
We shall also take a more algebraic geometric viewpoint then has historically been
done. Indeed the second two parts of the book, will be based on such a viewpoint. In our
characteristic free approach, this means a firmer focus on quadratic forms which have nice
geometric objects attached to them rather than bilinear forms. We do this for a variety
of reasons.
Firstly, one can associate to a quadratic form a number of algebraic varieties: the
quadric of isotropic lines in a projective space and, more generally, for an integer i > 0
the variety of isotropic subspaces of dimension i. More importantly, basic properties of
quadratic forms can be reformulated in terms of the associated varieties: a quadratic
form is isotropic if and only if the corresponding quadric has a rational point. A non-
degenerate quadratic form is hyperbolic if and only if the variety of maximal totally
isotropic subspaces has a rational point.
Not only are the associated varieties important but so are the morphisms between
them. Indeed if ϕ is a quadratic form over F and L/F a finitely generated field extension
then there is a variety Y over F with function field L, and the form ϕ is isotropic over L
if and only if there is a rational morphism from Y to the quadric of ϕ.
Working with correspondences rather than just rational morphisms adds further depth
to our study, where we identify morphisms with their graphs. Working with these leads to
the category of Chow correspondences. This provides greater flexibility, because we can
view correspondences as elements of Chow groups and apply the rich machinery of that
theory: pull-back and push-forward homomorphisms, Chern classes of vector bundles,
and Steenrod operations. For example, suppose we wish to prove that a property A of
quadratic forms implies a property B. We translate the properties A and B to “geometric”
properties A0 and B 0 about the existence of certain cycles on certain varieties. Starting
with cycles satisfying A0 we then can attempt to apply the operations over the cycles as
above to produce cycles satisfying B 0 .
All the varieties listed above are projective homogeneous varieties under the action
of the orthogonal group or special orthogonal group of ϕ, i.e., the orthogonal group acts
transitively on the varieties. It is not surprising that the properties of quadratic forms
INTRODUCTION ix
are reflected in the properties of the special orthogonal groups. For example, if ϕ is of
dimension 2n or 2n − 1 (with n ≥ 2) then the special orthogonal group is a semisimple
group of type Dn or Bn . The classification of semisimple groups is characteristic free. This
explains why most important properties of quadratic forms hold in all characteristics.
Unfortunately, bilinear forms are not “geometric”. We can associate varieties to a
bilinear form, but it would be a variety of the associated quadratic form. Moreover in
characteristic two the automorphism group of a bilinear form is not semisimple.
In this book we sometimes give several proofs of the same results – one is classical,
another is geometric. (This can be the same proof, but written in geometric language.)
For example, this is done for Springer’s theorem and the Separation Theorem.
The first part of the text will derive classical results under this new setting. It is
self-contained needing minimal prerequisites except for Chapter VII. In this chapter we
shall assume the results of Voevodsky in [142] and Orlov-Vishik-Voevodsky in [110] for
fields of characteristic not two and Kato in [76] for fields of characteristic two on the
solution for the analog of the Milnor Conjecture in algebraic K-theory. We do give new
proofs for the case n = 2.
Prerequisites for the second two parts of the text will be more formidable. A reasonable
background in algebraic geometry will be assumed. For the convenience of the reader
appendices have been included to aid the reader. Unfortunately, we cannot give details
of [142] or [110] as it would lead us from the methods at hand.
The first part of this book covers the “classical” theory of quadratic forms, i.e., without
heavy use of algebraic geometry, bringing it up to date. As the characteristic of a field is
not deemed to be different from two, this necessitates a bifurcation of the theory in the
theory of symmetric bilinear forms and the theory of quadratic forms. The introduction
of these subjects is given in the first two chapters.
Chapter I investigates the foundations of the theory of symmetric bilinear forms over
a field F . Two major consequences of dealing with arbitrary characteristic is that such
forms may not be diagonalizable and that non-degenerate isotropic planes need not be
hyperbolic. With this taken into account, standard Witt theory, to the extent possible,
is developed. In particular, Witt decomposition still holds, so that the Witt ring can be
constructed in the usual way as well as the classical group presentation of the Witt ring
W (F ). This presentation is generalized to the fundamental ideal I(F ) of even dimensional
forms in W (F ) and then to the the second power I 2 (F ) of I(F ), a theme returned to in
Chapter VII. The Stiefel-Whitney invariants of bilinear forms are introduced and its
relationship with the invariants ēn : I n (F )/I n+1 (F ) → Kn (F )/2Kn+1 (F ) for n = 1, 2.
The theory of bilinear Pfister forms is introduced and some basic properties developed.
Following [35], we introduce chain p-equivalence and linkage of Pfister forms as well as
introducing annihilators of Pfister forms in the Witt ring.
Chapter II investigates the foundations of the theory of quadratic forms over a field
F . Because of the arbitrary characteristic assumption on the field F , the definition of
non-degenerate must be made more carefully, and quadratics forms are far from having
orthogonal bases in general. Much of Witt theory however goes through as the Witt
Extension Theorem holds for quadratic forms under fairly weak assumptions hence Witt
x CONTENTS
Decomposition. The Witt group Iq (F ) of non-degenerate even dimensional quadratic

forms is defined and shown to be a W (F )-module. The theory of quadratic Pfister forms
is introduced and some results analogous to that of the bilinear case is introduced. More-
over, cohomological invariants of quadratic Pfister forms are introduced and some pre-
liminary results about them and their extension to the appropriate filtrant of the Witt
group of quadratic forms discussed. In addition the classical quadratic form invariants:
discriminant and Clifford invariant, are defined.
Chapter III begins the utilization of function field techniques in the study of quadratic
forms all done without restriction of characteristic. The classical Cassels-Pfister theorem
is established. Values of anisotropic quadratic and bilinear forms over a polynomial ring
are investigated, special cases being the representation of one form as a subform of another
and various norm principles due to Knebusch (cf. [82]). To investigate norm principles
of similarity factors due to Scharlau (cf. [123]), quadratic forms over valuation rings and
transfer maps are introduced.
Chapter IV introduces algebraic geometric methods, i.e., looking at the theory under
the base extension of the function field of a fixed quadratic form. In particular, the
notion of domination of one form by another is introduced where an anisotropic quadratic
form ϕ is said to dominate an anisotropic quadratic form ψ (both of dimension of at
least two) if ϕF (ψ) is isotropic. The geometric properties of Pfister forms are developed
leading to the Arason-Pfister Hauptsatz that non-zero anisotropic quadratic (respectively,
symmetric bilinear) forms in Iqn (F ) (respectively, I n (F )) are of dimension at least 2n and
its application to linkage of Pfister forms. Knebusch’s generic tower of an anisotropic
quadratic form is introduced and the W (F )-submodules Jn (F ) of Iq (F ) defined by the
notion of degree. These submodules turn out to be precisely the corresponding Iqn (F )
(to be shown in Chapter VII). Hoffman’s Separation Theorem that if ϕ and ψ are two
anisotropic quadratic forms over F with dim ϕ ≤ 2n < dim ψ for some n ≥ 0 then
ϕF (ψ) is anisotropic is proven as well as Fitzgerald’s theorem characterizing quadratic
Pfister forms. In addition, excellent forms and extensions are discussed. In particular,
Arason’s result that the extension of a field by the function field of a non-degenerate three
dimensional quadratic form is excellent is proven. The chapter ends with discussion of
central simple algebras over the function field of a quadric.
Chapter V studies symmetric bilinear and quadratic forms under field extensions. The
chapter begins with the study of the structure of the Witt ring of a field F based on the
work of Pfister. After dispensing with the non-formally real F , we turn to the study over
a formally real field utilizing the theory of pythagorean fields and the pythagorean closure
of a field leading to the Local-Global Theorem of Pfister and its consequences to structure
of the Witt ring over a formally real field. The total signature map from the Witt ring
to the ring of continuous functions from the order space of a field to the integers is then
carefully studied, in particular, the approximation of elements in this ring of functions
by quadratic forms. The behavior of quadratic and bilinear forms under a quadratic
extension (both separable and inseparable) is then investigated. The special case of the
torsion of the Witt ring under such extensions is studied. A detailed investigation of
torsion Pfister forms is begun, leading to the theorem of Krüskemper that implies if K/F
is a quadratic field extension with I n (K) = 0 then I n (F ) is torsion-free.
INTRODUCTION xi
Chapter VI studies u-invariants, their behavior under field extensions, and values that
they can take. Special attention is given to the case of formally real fields.
Chapter VII establishes consequences of the result of Orlov-Vishik-Voevodsky in [110]
which we assume in this chapter. In particular, answers and generalizations of results from
the previous chapters are established. For fields of characteristic not two, the ideals I n (F )
and Jn (F ) are shown to be identical. The annihilators of Pfister forms in the Witt ring
are shown to filter through the I n (F ), i.e., the intersection of such annihilators and I n (F )
are generated by Pfister forms in the intersection. A consequence is that torsion in I n (F )
is generated by torsion n-fold Pfister forms solving a conjecture of Lam. A presentation
for the group structure of the I n (F )’s is determined generalizing that given for I 2 (F ) in
Chapter I. Finally, it is shown if K/F is a finite generated field extension of transcendental
degree m then I n (K) torsion free implies the same for I n−m (F ).
In Chapter VIII, we give a new elementary proof of the theorem in [101] that the
second cohomological invariant is an isomorphism in the case that the characteristic of
the field is different from two. (The case of characteristic two having been done in Chapter
II.) This is equivalent to the degree two case of the Milnor Conjecture in [109] stating
that the norm residue homomorphism
hnF : Kn (F )/2Kn (F ) → H n (F, Z/2Z)
is an isomorphism for every integer n. The Milnor Conjecture was proven in full by
V. Voevodsky in [142]. Unfortunately, the scope of this book does not allow us to prove
this beautiful result as the proof requires motivic cohomology and Steenrod operations
developed by Voevodsky. In Chapter VIII, we give an “elementary” proof of the degree
two case of the Milnor Conjecture that does not rely either on a specialization argument
or on higher K-theory as did the original proof of this case in [102].
In the second part of the book, we develop the needed tools in algebraic geometry that
will be applied in the third part. The main object studied in Part Two is the Chow group
of algebraic cycles modulo rational equivalence on an algebraic scheme. Using algebraic
cycles, we introduce the category of correspondences.
In Chapter IX (following the approach of [119] given by Rost), we develop the K-
homology and K-cohomology theories of schemes over a field. This generalizes the Chow
groups. We establish functorial properties of these theories (pull-back, push-forward,
deformation and Gysin homomorphisms), introduce Euler and Chern classes of vector
bundles and prove basic results such as the Homotopy Invariance and Projective Bundle
Theorems. We apply these results to Chow groups in the next chapter.
Chapter XI is devoted to the study of Steenrod operations on Chow groups modulo 2
over fields of characteristic not two. Steenrod operations for motivic cohomology modulo
a prime integer p of a scheme X were originally constructed by Voevodsky in [144]. The
reduced power operations (but not the Bockstein operation) restrict to the Chow groups
of X. An “elementary” construction of the reduced power operations modulo p on Chow
groups (requiring equivariant Chow groups) was given by Brosnan in [22].
In Chapter XII, we introduced the notion of a Chow motive that is due to A. Grothendieck.
Many (co)homology theories defined on the category of smooth complete varieties, such as
Chow groups and more generally the K-(co)homology groups take values in the category
xii CONTENTS
of abelian groups. But the category of smooth complete varieties itself does not have the
structure of an additive category as we cannot add morphisms of varieties. The category
of Chow motives, however, is an additive tensor category. This additional structure gives
more flexibility when working with regular and rational morphisms.
In the third part of the book we apply algebraic geometric methods to the further
study of quadratic forms. In Chapter XIII, we prove preliminary facts about algebraic
cycles on quadrics and their powers. We also introduce the basic combinatorial object
associated to a quadratic form – shell triangles and diagrams of cycles. The corresponding
pictures of these shell triangles simplify visualization of algebraic cycles and operations
over the cycles.
In Chapter XIV, we study the Izhboldin dimension of smooth projective quadrics. It
is defined as the integer
dimIzh (X) := dim X − i1 (X) + 1 ,
where i1 (X) is the first Witt index of the quadric X. The Izhboldin dimension behaves
better than the classical dimension with respect to splitting properties. For example, if
X and Y are anisotropic smooth projective quadrics and Y is isotropic over the function
field F (X) then dimIzh X ≤ dimIzh Y but dim X may be bigger than dim Y .
Chapter XV is devoted to applications of the Steenrod operations. The following
problems are solved:
(1) All possible values of the first Witt index of a quadratic form are determined.
(2) All possible values of dimensions of anisotropic quadratic forms in I n (F ) are
determined.
(3) It is shown that excellent forms have the smallest height among all quadratic
forms of given dimension.
In Chapter XVI, we study the variety of maximal isotropic subspaces of a quadratic
forms. A discrete invariant J(ϕ) of a quadratic form ϕ is introduced. We also introduce
the notion of canonical dimension and compute it for projective quadrics and varieties of
totally isotropic subspaces.
In the last chapter we study motives of smooth projective quadrics in the category of
correspondences and motives.
This book could not have been written without the help and encouragement of many
of our friends, collaborators, and students nor the many researchers whose work was es-
sential to producing this volume. These are too numerous to be listed here individually.
We do wish to mention those who gave us valuable advise in the preparation and writing
of this tome and those who helped proofread the manuscript: Jón Arason, Ricardo Baeza,
Alex Boisvert, Detlev Hoffmann, Bryant Mathews, Claus Schubert, Jean-Pierre Tignol,
Alexander Vishik. In addition, we would like to thank the referees who made valuable
suggestions to improve the manuscript and Sergei Gelfand and the American Mathemat-
ical Society for its encouragement. It also gives us great pleasure to thank the National
Science Foundation for its generous support. Finally, we wish to thank our families who
put up with us through the long process of bringing this volume to fruition.
Part
Classical theory of symmetric bilinear forms

and quadratic forms
CHAPTER I
Bilinear Forms
1. Foundations
The study of (n × n)-matrices over a field F leads to various classification problems.

Of special interest is to classify alternating and symmetric matrices. If A and B are
two such matrices, we say that they are congruent if A = P t BP for some invertible
matrix P . For example, it is well-known that symmetric matrices are diagonalizable if
the characteristic of F is different from two. So the problem of classifying matrices up to
congruence reduces to the study of a congruence class of a matrix in this case. The study
of alternating and symmetric bilinear forms over an arbitrary field is the study of this
problem in a coordinate-free approach. Moreover, we shall, whenever possible, give proofs
independent of characteristic. In this section, we introduce the definitions and notations
needed throughout the text and prove that we have a Witt Decomposition Theorem (cf.
Theorem 1.28 below) for such forms. As we make no assumption on the characteristic of
the underlying field, this makes the form of this theorem more delicate.
Definition 1.1. Let V be a finite dimensional vector space over a field F . A bilinear
form on V is a map b : V × V → F satisfying for all v, v 0 , w, w0 ∈ V and c ∈ F
b(v + v 0 , w) = b(v, w) + b(v 0 , w)
b(v, w + w0 ) = b(v, w) + b(v, w0 )
b(cv, w) = cb(v, w) = b(v, cw).
The bilinear form is called symmetric if b(v, w) = b(w, v) for all v, w ∈ V and is called
alternating if b(v, v) = 0 for all v ∈ V . If b is an alternating form, expanding b(v+w, v+w)
shows that b is skew symmetric, i.e., that b(v, w) = −b(w, v) for all v, w ∈ V . In
particular, every alternating form is symmetric if char F = 2. We call dim V the dimension
of the bilinear form and also write it as dim b. We write b is a bilinear form over F if b
is a bilinear form on a finite dimensional vector space over F and denote the underlying
space by Vb .
Definition 1.2. Let V ∗ := HomF (V, F ) denote the dual space of V . A bilinear form
b on V is called non-degenerate if l : V → V ∗ defined by v 7→ lv : w 7→ b(v, w) is an
isomorphism. An isometry f : b1 → b2 between two¡ bilinear forms
¢ bi , i = 1, 2, is a linear
isomorphism f : Vb1 → Vb2 such that b1 (v, w) = b2 f (v), f (w) for all v, w ∈ Vb1 . If such
an isometry exists, we write b1 ' b2 and say that b1 and b2 are isometric.
Let b be a bilinear
¡ form
¢ on V . Let {v1 , . . . , vn } be a basis for V . Then¡b is determined
¢
by the matrix b(vi , vj ) and the form is non-degenerate if and only if b(vi , vj ) is in-
vertible. Conversely any matrix B in the n × n matrix ring Mn (F ) determines a bilinear
3
4 I. BILINEAR FORMS
form based on V . If b is symmetric (respectively, alternating) then the associated matrix

is symmetric (respectively, alternating where a square matrix (aij ) is called alternating if
aij = −aji and aii = 0 for all i, j). Let b and b0 be two bilinear forms with matrices B
and B 0 relative to some bases. Then b ' b0 if and only if B 0 = At BA for some invertible
matrix A, i.e., the matrices B 0 and B are congruent. As det B 0 = det B · (det A)2 and
det A 6= 0, the determinant of B 0 coincides with the determinant of B up to squares. We
define the determinant of a non-degenerate bilinear form b by det b := det B · F ×2 in
F × /F ×2 , where B is a matrix representation of b and F × is the multiplicative group of
F (and more generally, R× denotes the unit group of a ring R). So the det is an invariant
of the isometry class of a non-degenerate bilinear form.
The set Bil(V ) of bilinear forms on V is a vector space over F . The space Bil(V )
contains the subspaces Alt(V ) of alternating forms on V and Sym(V ) of symmetric bilinear
forms on V . The correspondence of bilinear forms and matrices given above defines a linear
isomorphism Bil(V ) → Mdim V (F ). If b ∈ Bil(V ) then b − bt is alternating where the
bilinear form bt is defined by bt (v, w) = b(w, v) for all v, w ∈ V . Since every alternating
n × n-matrix is of the form B − B t for some B, the linear map Bil(V ) → Alt(V ) given
by b 7→ b − bt is surjective. Therefore, we have an exact sequence of vector spaces
(1.3) 0 → Sym(V ) → Bil(V ) → Alt(V ) → 0.
Exercise 1.4. Construct natural isomorphisms
V2 V2
Bil(V ) ' (V ⊗F V )∗ ' V ∗ ⊗F V ∗ , Sym(V ) ' S 2 (V )∗ , Alt(V ) ' (V )∗ ' (V ∗ )
and show that the exact sequence 1.3 is dual to the standard exact sequence
V
0 → 2 (V ) → V ⊗F V → S 2 (V ) → 0.
V
where 2 (V ) is the exterior square of V and S 2 (V ) is the symmetric square of V .
If b, c ∈ Bil(V ), we say the two bilinear forms b and c are similar if b ' ac for some
a ∈ F ×.
Let V be a finite dimensional vector space over F and let λ = ±1. Define the hyperbolic
λ-bilinear form on V to be Hλ (V ) = bHλ on V ⊕ V ∗ with
bHλ (v1 + f1 , v2 + f2 ) := f1 (v2 ) + λf2 (v1 )
for all v1 , v2 ∈ V and f1 , f2 ∈ V ∗ . If λ = 1, the form Hλ (V ) is a symmetric bilinear
form and if λ = −1, it is an alternating bilinear form. A bilinear form b is called a
hyperbolic bilinear form if b ' Hλ (W ) for some finite dimensional F -vector space W and
some λ = ±1. The hyperbolic form Hλ (F ) is called the hyperbolic plane and denoted Hλ
. It has the matrix representation µ ¶
0 1
λ 0
in the appropriate basis. If b ' Hλ , then b has the above matrix representation in some
basis {e, f } of Vb . We call e, f a hyperbolic pair. Hyperbolic forms are non-degenerate.
Let b be a bilinear form on V and W ⊂ V a subspace. The restriction of b to W is a
bilinear form on W and is called a subform of b. We denote this form by b|W .
1. FOUNDATIONS 5
1.A. Structure theorems for bilinear forms. Let b be a symmetric or alternating

bilinear form on V . We say v, w ∈ V are orthogonal if b(v, w) = 0. Let W, U ⊂ V be
subspaces. Define the orthogonal complement of W by
W ⊥ := {v ∈ V | b(v, w) = 0 for all w ∈ W }.
This is a subspace of V . We say W is orthogonal to U if W ⊂ U ⊥ , equivalently U ⊂ W ⊥ .
If V = W ⊕ U is a direct sum of subspaces with W ⊂ U ⊥ , we write b = b|W ⊥ b|U and
say b is the the (internal) orthogonal sum of b|W and b|U . The subspace V ⊥ is called the
radical of b and denoted by rad b . The form b is non-degenerate if and only if rad b = 0.
If K/F is a field extension, let VK := K ⊗F V , a vector space over K. We have the
standard embedding V → VK by v 7→ 1 ⊗ v. Let bK denote the extension of b to VK , so
bK (a ⊗ v, c ⊗ w) = acb(v, w) for all a, c ∈ K and v, w ∈ V . The form bK is of the same
type as b. Moreover, rad bK = (rad b)K hence b is non-degenerate if and only if bK is
non-degenerate.
Let : V → V / rad b be the canonical epimorphism. Define b to be the bilinear form
on V determined by b(v1 , v2 ) := b(v1 , v2 ) for all v1 , v2 ∈ V . Then b is a non-degenerate
bilinear form of the same type as b. Note also that if f : b1 → b2 is an isometry of
symmetric or alternative bilinear forms then f (rad b1 ) = rad b2 .
We have
Lemma 1.5. Let b be a symmetric or alternating bilinear form on V . Let W be any
subspace of V such that V = rad b ⊕ W . Then b|W is non-degenerate and
b = b|rad b ⊥ b|W = 0|rad b ⊥ b|W
with b|W ' b, the form induced on V / rad b. In particular, b|W is unique up to isometry.
The lemma above shows that it is sufficient to classify non-degenerate bilinear forms.
In general, if b is a symmetric or alternating bilinear form on V and W ⊂ V is a subspace
then we have an exact sequence of vector spaces
l
0 → W ⊥ → V −W
→ W ∗,
where lW is defined by v 7→ lv |W : x 7→ b(v, x). Hence dim W ⊥ ≥ dim V − dim W . It is
easy to determine when this is an equality.
Proposition 1.6. Let b be a symmetric or alternating bilinear form on V . Let W be
any subspace of V . Then the following are equivalent:
(1) W ∩ rad b = 0.
(2) lW : V → W ∗ is surjective.
(3) dim W ⊥ = dim V − dim W .
∗
Proof. (1) holds if and only if the map lW : W → V ∗ is injective if and only if the
∗
map lW : V → W is surjective if and only if (3) holds. ¤
Note that the conditions (1) − (3) hold if either b or b|W is non-degenerate.
A key observation is
6 I. BILINEAR FORMS
Proposition 1.7. Let b be a symmetric or alternating bilinear form on V . Let W be

a subspace such that b|W is non-degenerate. Then b = b|W ⊥ b|W ⊥ . In particular, if b is
also non-degenerate so is b|W ⊥ .
Proof. By Proposition 1.6, dim W ⊥ = dim V − dim W hence V = W ⊕ W ⊥ . The
result follows. ¤
Corollary 1.8. Let b be a symmetric bilinear form on V . Let v ∈ V satisfy b(v, v) 6=
0. Then b = b|F v ⊥ b(F v)⊥ .
Let b1 and b2 be two symmetric or alternating bilinear forms on V1 and V2 respectively.
`
Then their external orthogonal sum b, denoted by b1 ⊥ b2 , is the form on V1 V2 given
by ¡ ¢
b (v1 , v2 ), (w1 , w2 ) := b1 (v1 , w1 ) + b2 (v2 , w2 )
for all vi , wi ∈ Vi , i = 1, 2.
If n is a non-negative integer and b is a symmetric or alternating bilinear form over
F , abusing notation, we let
nb := b | ⊥ ·{z
· · ⊥ b} .
n
In particular, if n is a non-negative integer, we do not interpret nb with n viewed in the
field.
For example, Hλ (V ) ' nHλ for any n-dimensional vector space V over F .
It is now easy to complete the classification of alternating forms.
Proposition 1.9. Let b be a non-degenerate alternating form on V . Then dim V = 2n
for some n and b ' nH−1 , i.e., b is hyperbolic.
Proof. Let 0 6= v ∈ V . Then there exists w ∈ V such that b(v, w) = a 6= 0.
Replacing w by a−1 w, we see that v, w is a hyperbolic pair in the space W = F v ⊕ F w,
so b|W is a hyperbolic subform of b, in particular, non-degenerate. Therefore, b = b|W ⊥
b|W ⊥ by Proposition 1.7. The result follows by induction on dim b. ¤
The proof shows that every non-degenerate alternating form b on V has a symplectic
basis, i.e., a basis {v1 , . . . , v2n } for V satisfying b(vi , vn+i ) = 1 for all i = 1, . . . , n and
b(vi , vj ) = 0 if i ≤ j and j 6= n + i.
We turn to the classification of the isometry type of symmetric bilinear forms. By
Lemma 1.5, Corollary 1.8, and induction, we therefore have the following
Corollary 1.10. Let b be a symmetric bilinear form on V . Then
b = b|rad b ⊥ b|V1 ⊥ · · · ⊥ b|Vn ⊥ b|W
with Vi a one-dimensional subspace of V and b|Vi non-degenerate for all i = 1, . . . , n and
b|W a non-degenerate alternating subform on a subspace W of V .
If char F 6= 2 then, in the corollary, b|W is symmetric and alternating hence W = {0}.
In particular, every bilinear form b has an orthogonal basis, i.e., a basis {v1 , . . . , vn } for
Vb satisfying b(vi , vj ) = 0 if i 6= j. The form is non-degenerate if and only if b(vi , vi ) 6= 0
for all i.
1. FOUNDATIONS 7
If char F = 2, by Proposition 1.9, the alternating form b|W in the corollary above has
a symplectic basis and satisfies b|W ' nH1 for some n.
Let a ∈ F . Denote the bilinear form on F given by b(v, w) = avw for all v, w ∈ F by
haib or simply hai. In particular, hai ' hbi if and only if a = b = 0 or aF ×2 = bF ×2 in
F × /F ×2 . Denote
ha1 i ⊥ · · · ⊥ han i by ha1 , . . . , an ib or simply by ha1 , . . . , an i.
We call such a form a diagonal form. A symmetric bilinear form b isometric to a diagonal
form is called diagonalizable. Consequently, b is diagonalizable if and only if b has an
orthogonal basis. Note that detha1 , . . . , an i = a1 · · · an F ×2 if ai ∈ F × for all i. Corollary
1.10 says that every bilinear form b on V satisfies
b ' rh0i ⊥ ha1 , . . . , an i ⊥ b0
with r = dim(rad b) and b0 an alternating form and ai ∈ F × for all i. In particular, if
char F 6= 2 then every symmetric bilinear form is diagonalizable.
Example 1.11. Let a, b ∈ F × . Then h1, ai ' h1, bi if and only if aF ×2 = deth1, ai =
deth1, bi = bF ×2 .
1.B. Values and similarities of bilinear forms. We study the values that a bi-
linear form can take as well as the similarity factors. We begin with some notation.
Definition 1.12. Let b be a bilinear form on V over F . Let
D(b) := {b(v, v) | v ∈ V with b(v, v) 6= 0},
the set on nonzero values of b and
G(b) := {a ∈ F × | ab ' b},
a group called the group of similarity factors of b . Also set
e
D(b) := D(b) ∪ {0}.
e
We say that elements in D(b) are represented by b.
For example, G(H1 ) = F × . A symmetric bilinear form is called round if G(b) = D(b). In
particular, if b is round then D(b) is a group.
Remark 1.13. If b is a symmetric bilinear form and a ∈ D(b) then b ' hai ⊥ c for
some symmetric bilinear form c by Corollary 1.8.
Lemma 1.14. Let b be a bilinear form. Then
D(b) · G(b) ⊂ D(b).
In particular, if 1 ∈ D(b) then G(b) ⊂ D(b).
Proof. Let a ∈¡ G(b) and ¢b ∈ D(b). Let λ : b → ab be an isometry and v ∈ Vb satisfy
b = b(v, v). Then b λ(v), λ(v) = ab(v, v) = ab. ¤
8 I. BILINEAR FORMS
Example 1.15. Let K = F [t]/(t2 − a) with a ∈ F , where F [t] is the polynomial

ring over F . So K = F ⊕ F θ as a vector space over F where θ denotes the class of t
in K. If z = x + yθ with x, y ∈ F , write z = x − yθ. Let s : K → F be the F -linear
functional defined by s(x + yθ) = x. Then b defined by b(z1 , z2 ) = s(z1 z 2 ) is a binary
symmetric bilinear form on K. Let N (z) = zz for z ∈ K. Then D(b) = {N (z) 6= 0 | z ∈
K} = {N (z) | z ∈ K × }. If z ∈ K then λz : K → K given by w → zw is an F -linear
isomorphism if and only if N (z) 6= 0. Suppose that λz is an F -isomorphism. As
b(λz z1 , λz z2 ) = b(zz1 , zz2 ) = N (z)s(z1 z 2 ) = N (z)b(z1 , z2 ),
we have an isometry N (z)b ' b for all z ∈ K × . In particular, b is round. Computing b
on the orthogonal basis {1, θ} for K shows that b is isometric to the bilinear form h1, −ai.
If a ∈ F × then b ' h1, −ai is non-degenerate.
Remark 1.16. (i) Let b be a binary symmetric bilinear form on V . Suppose there
exists a basis {v, w} for V satisfying b(v, v) = 0, b(v, w) = 1, and b(w, w) = a 6= 0. Then
b is non-degenerate as the matrix corresponding to b in this basis is invertible. Moreover,
{w, −av + w} is an orthogonal basis for V and, using this basis, we see that b ' ha, −ai.
(ii) Suppose that char F 6= 2. Let b = ha, −ai with a ∈ F × and {e, g} an orthogonal
1
basis for Vb satisfying a = b(e, e) = −b(f, f ). Evaluating on the basis {e + f, 2a (e − f )}
shows that b ' H1 . In particular, ha, −ai ' H1 for all a ∈ F . Moreover, ha, −ai ' H1 is
×
round and universal, where a non-degenerate symmetric bilinear form b is called universal
if D(b) = F × .
(iii) Suppose that char F = 2. As H1 = H−1 is alternating while ha, ai ¡ is not, ¡ 6'¢H1
¢ ha, ai
×
for any a ∈ F . ¡Moreover,¢ H1 is not round since D(H1 ) = ∅. As D ha, ai = D hai =
aF , we have G ha, ai = F ×2 by Lemma 1.14. In particular, ha, ai is round if and only
×2
if a ∈ F ×2 and ha, ai ' hb, bi if and only if aF ×2 ' bF ×2 .

(iv) Witt Cancellation holds if char F 6= 2, i.e., if there exists an isometry of symmetric
bilinear forms b ⊥ b0 ' b ⊥ b00 over F with b non-degenerate then b0 ' b00 . (Cf. Theorem
8.4 below.) If char F = 2, this is false in general. For example,
h1, 1, −1i ' h1i ⊥ H1
over any field. Indeed if b is three dimensional on V and V has an orthogonal basis
{e, f, g} with b(e, e) = 1 = b(f, f ) and b(g, g) = −1 then the right hand side arises from
the basis {e + f + g, e + g, −f − g}. But by (iii), h1, −1i 6' H1 if char F = 2. Multiplying
the equation above by any a ∈ F × , we also have
(1.17) ha, a, −ai ' hai ⊥ H1 .
Proposition 1.18. Let b be a symmetric bilinear form. If D(b) 6= ∅ then b is diag-
onalizable. In particular, a nonzero symmetric bilinear form is diagonalizable if and only
if it is not alternating.
Proof. If a ∈ D(b) then
b ' hai ⊥ b1 ' hai ⊥ rad b1 ⊥ c1 ⊥ c2
with b1 a symmetric bilinear form by Corollary 1.8 and c1 a non-degenerate diagonal form
and c2 a non-degenerate alternating form by Corollary 1.10. By the remarks following
1. FOUNDATIONS 9
Corollary 1.10, we have c2 = 0 if char F 6= 2 and c2 = mH1 for some integer m if

char F = 2. By 1.17, we conclude that b is diagonalizable in either case.
If b is not alternating then D(b) 6= ∅ hence b is diagonalizable. Conversely, if b is
diagonalizable, it cannot be alternating as it is not the zero form. ¤
Corollary 1.19. Let b be a symmetric bilinear form over F . Then b ⊥ h1i is
diagonalizable.
Let b be a symmetric bilinear form on V . A vector v ∈ V is called anisotropic if
b(v, v) 6= 0 and isotropic if v 6= 0 and b(v, v) = 0. We call b anisotropic if there are no
isotropic vectors in V and isotropic otherwise.
Corollary 1.20. Every anisotropic bilinear form is diagonalizable.
Note that an anisotropic symmetric bilinear form is non-degenerate as its radical is
trivial.
Example 1.21. Let F be a quadratically closed field, i.e., every element in F is a
square. Then, up to isometry, 0 and h1i are the only anisotropic forms over F . In
particular, this applies if F is algebraically closed.
An anisotropic form may not be anisotropic under base extension. However, we do
have:
Lemma 1.22. Let b be an anisotropic bilinear form over F . If K/F is purely tran-
scendental then bK is anisotropic.
Proof. First suppose that K = F (t), the field of rational functions over F in the
variable t. Suppose that bF (t) is isotropic. Then there exist a vector 0 6= v ∈ VbF (t) such
that bF (t) (v, v) = 0. Multiplying by an appropriate nonzero polynomial, we may assume
that v ∈ F [t] ⊗F V . Write v = v0 + t ⊗ v1 + · · · + tn ⊗ vn with v1 , . . . , vn ∈ V and vn 6= 0.
As the t2n coefficient b(vn , vn ) of b(v, v) must vanish, vn is an isotropic vector of b, a
contradiction.
If K/F is finitely generated then the result follows by induction on the transcendence
degree of K over F . In the general case, if bK is isotropic there exists a finitely generated
purely transcendental extension K0 of F in K with bK0 isotropic, a contradiction. ¤
1.C. Metabolic bilinear forms. Let b be a symmetric bilinear form on V . A

subspace W ⊂ V is called a totally isotropic subspace of b if b|W = 0, i.e., if W ⊂ W ⊥ .
If b is isotropic then it has a nonzero totally isotropic subspace. Suppose that b is non-
degenerate and W is a totally isotropic subspace. Then dim W + dim W ⊥ = dim V by
Proposition 1.6 hence dim W ≤ 21 dim V . We say that W is a lagrangian for b if we have an
equality dim W = 12 dim V , equivalently W ⊥ = W . A non-degenerate symmetric bilinear
form is called metabolic if it has a lagrangian. In particular, every metabolic form is even
dimensional. Clearly an orthogonal sum of metabolic forms is metabolic.
Example 1.23. (1) Symmetric hyperbolic forms are metabolic.
(2) The form b ⊥ (−b) is metabolic if b is any non-degenerate symmetric bilinear form.
10 I. BILINEAR FORMS
(3) A 2-dimensional metabolic space is nothing but a non-degenerate isotropic plane.

A metabolic plane is therefore either isomorphic to ha, −ai for some a ∈ F × or to the
hyperbolic plane H1 by Remark 1.16. In particular, the determinant of a metabolic plane
is −F ×2 . If char F 6= 2 then ha, −ai ' H1 by Remark 1.16, so in this case, every metabolic
plane is hyperbolic.
Lemma 1.24. Let b be an isotropic non-degenerate symmetric bilinear form over V .
Then every isotropic vector belongs to a 2-dimensional metabolic subform.
Proof. Suppose that b(v, v) = 0 with v 6= 0. As b is non-degenerate, there exists a
u ∈ V such that b(u, v) 6= 0. Then b|F v⊕F u is metabolic. ¤
Corollary 1.25. Every metabolic form is an orthogonal sum of metabolic planes. In
dim b
particular, if b is a metabolic form over F then det b = (−1) 2 F ×2 .
Proof. We induct on the dimension of a metabolic form b. Let W ⊂ V = Vb be
a lagrangian. By Lemma 1.24, a nonzero vector v ∈ W belongs to a metabolic plane
P ⊂ V . It follows from Proposition 1.7 that b = b|P ⊥ b|P ⊥ and by dimension count that
W ∩ P ⊥ is a lagrangian of b|P ⊥ as W cannot lie in P ⊥ . By the induction hypothesis, b|P ⊥
is an orthogonal sum of metabolic planes. The second statement follows from Example
1.23(3). ¤
Corollary 1.26. If char F 6= 2, the classes of metabolic and hyperbolic forms coin-
cide. In particular, every isotropic non-degenerate symmetric bilinear form is universal.
Proof. This follows from Remark 1.16 (ii) and Lemma 1.24 . ¤
Lemma 1.27. Let b and b0 be two symmetric bilinear forms. If b ⊥ b0 and b0 are both
metabolic so is b.
Proof. By Corollary 1.25, we may assume that b0 is 2-dimensional. Let W be a
lagrangian for b ⊥ b0 . Let p : W → Vb0 be the projection and W0 = ker(p) = W ∩ Vb .
Suppose that p is not surjective. Then dim W0 ≥ dim W − 1 hence W0 is a lagrangian of
b and b is metabolic.
So we may assume that p is surjective. Then dim W0 = dim W − 2. As b0 is metabolic,
it is isotropic. Choose an isotropic vector v 0 ∈ Vb0 and a vector w ∈ W such that p(w) = v 0 ,
i.e., w = v + v 0 for some v ∈ Vb . In particular, b(v, v) = (b ⊥ b0 )(w, w) − b0 (v 0 , v 0 ) = 0.
Since W0 ⊂ Vb , we have v 0 is orthogonal to W0 hence v is also orthogonal to W0 . If we
show that v 0 6∈ W then v ∈ / W0 and W0 ⊕ F v is a lagrangian of b and b is metabolic.
So suppose v ∈ W . There exists v 00 ∈ Vb0 such that b0 (v 0 , v 00 ) 6= 0 as b0 is non-
0
degenerate. Since p is surjective, there exists w00 ∈ W with w00 = u00 + v 00 for some
u00 ∈ Vb . As W is totally isotropic,
0 = (b ⊥ b0 )(v 0 , w00 ) = (b ⊥ b0 )(v 0 , u00 + v 00 ) = b0 (v 0 , v 00 ),
a contradiction. ¤
1.D. Witt Theory. We have the following form of the classical Witt Decomposition
Theorem (cf. [146]) for symmetric bilinear forms over a field of arbitrary characteristic.
2. THE WITT AND WITT-GROTHENDIECK RINGS OF SYMMETRIC BILINEAR FORMS 11
Theorem 1.28. (Bilinear Witt Decomposition Theorem) Let b be a non-degenerate

symmetric bilinear form on V . Then there exist subspaces V1 and V2 of V such that
b = b|V1 ⊥ b|V2 with b|V1 anisotropic and b|V2 metabolic. Moreover, b|V1 is unique up to
isometry.
Proof. We prove existence of the decomposition by induction on dim b. If b is
isotropic, there is a metabolic plane P ⊂ V by Lemma 1.24. As b = b|P ⊥ b|P ⊥ , the
proof of existence follows by applying the induction hypothesis to b|P ⊥ .
To prove uniqueness, assume that b1 ⊥ b2 ' b01 ⊥ b02 with b1 and b01 both anisotropic
and b2 and b02 both metabolic. We show that b1 ' b01 . The form
b1 ⊥ (−b01 ) ⊥ b2 ' b01 ⊥ (−b01 ) ⊥ b02
is metabolic, hence b1 ⊥ (−b01 ) is metabolic by Lemma 1.27. Let W be a lagrangian
of b1 ⊥ (−b01 ). Since b1 is anisotropic, the intersection W ∩ Vb1 is trivial. Therefore,
the projection W → Vb01 is injective and dim W ≤ dim b01 . Similarly, dim W ≤ dim b1 .
Consequently, dim b1 = dim W = dim b01 and the projections p : W → Vb1 and p0 : W →
Vb01 are isomorphisms. Let w = v + v 0 ∈ W , where v ∈ Vb1 and v 0 ∈ Vb01 . As
¡ ¢
0 = b1 ⊥ (−b01 ) (w, w) = b1 (v, v) − b01 (v 0 , v 0 ),
the isomorphism p0 ◦ p−1 : Vb1 → Vb01 is an isometry between b1 and b01 . ¤
Let b = b|V1 ⊥ b|V2 be the decomposition of the non-degenerate symmetric bilinear
form b on V in the theorem. The anisotropic form b|V1 , unique up to isometry, will be
denote by ban and called the anisotropic part of b. Note that the metabolic form b|V2 in
Theorem 1.28 is not unique in general by Remark 1.16 (iv). However, its dimension is
unique and even. Define the Witt index of b to be i(b) := (dim V2 )/2 .
Remark 1.16 (iv) also showed that the Witt Cancellation Theorem does not hold for
non-degenerate symmetric bilinear forms in characteristic two. The obstruction is the
metabolic forms. We have, however, the following
Corollary 1.29. (Witt Cancellation) Let b, b1 , b2 be non-degenerate symmetric
bilinear forms satisfying b1 ⊥ b ' b2 ⊥ b. If b1 and b2 are anisotropic then b1 ' b2 .
Proof. We have b1 ⊥ b ⊥ (−b) ' b2 ⊥ b ⊥ (−b) with b ⊥ (−b) metabolic. By
Theorem 1.28, b1 ' b2 . ¤
2. The Witt and Witt-Grothendieck rings of symmetric bilinear forms
In this section, we construct the Witt ring. The orthogonal sum induces an additive
structure on the isometry classes of symmetric bilinear forms. Defining the tensor product
of symmetric bilinear forms (corresponding to the classical Kronecker product of matrices)
turns this set of isometry classes into a semi-ring. The Witt Decomposition Theorem leads
to a nice description of the Grothendieck ring in terms of isometry classes of anisotropic
symmetric bilinear forms. The Witt ring W (F ) is the quotient of this ring by the ideal
generated by the hyperbolic plane.
Let b1 and b2 be symmetric bilinear forms over F . The tensor product of b1 and b2 is
defined to be the symmetric bilinear form b := b1 ⊗ b2 with underlying space Vb1 ⊗F Vb2
with the form b defined by
¡ ¢
b (v1 ⊗ v2 ), (w1 ⊗ w2 ) = b1 (v1 , w1 ) · b2 (v2 , w2 )
for all v1 , w1 ∈ Vb1 and v2 , w2 ∈ Vb2 . For example, if a ∈ F then hai ⊗ b1 ' ab1 .
Lemma 2.1. Let b1 and b2 be two non-degenerate bilinear forms over F . Then
(1) b1 ⊥ b2 is non-degenerate.
(2) b1 ⊗ b2 is non-degenerate.
(3) H1 (V ) ⊗ b1 is hyperbolic for all finite dimensional vector spaces V .
Proof. (1), (2): Let Vi = Vbi for i = 1, 2. The bi induce isomorphisms li : Vi → Vi∗
for i = 1, 2 hence b1 ⊥ b2 and b1 ⊗ b2 induce isomorphisms l1 ⊕ l2 : V1 ⊕ V2 → (V1 ⊕ V2 )∗
and l1 ⊗ l2 : V1 ⊗F V2 → (V1 ⊗F V2 )∗ respectively.
(3): Let {e, f } be a hyperbolic pair for H1 . Then the linear map (F ⊕ F ∗ ) ⊗F V1 →
V1 ⊕ V1∗ induced by e ⊗ v 7→ v and f ⊗ v 7→ lv : w 7→ b(w, v) is an isomorphism and
induces the isometry H1 ⊗ b → H1 (V ). ¤
It follows that the isometry classes of non-degenerate symmetric bilinear forms over
F is a semi-ring under orthogonal sum and tensor product. The Grothendieck ring of
this semi-ring is called the Witt-Grothendieck ring of F and denoted by W c (F ) . (Cf.
Scharlau [125] or Lang [91] for the definition and construction of the Grothendieck group
and ring.) In particular, every element in W c (F ) is a difference of two isometry classes
of non-degenerate symmetric bilinear forms over F . If b is a non-degenerate symmetric
c (F ). Thus if α ∈ W
bilinear form over F , we shall also write b for the class in W c (F ), there
exist non-degenerate symmetric bilinear forms b1 and b2 over F such that α = b1 − b2 in
c (F ). By definition, we have
W
c (F )
b1 − b2 = b01 − b02 in W
if and only if there exists a non-degenerate symmetric bilinear form b00 over F such that
(2.2) b1 ⊥ b02 ⊥ b00 ' b01 ⊥ b2 ⊥ b00 .
As any hyperbolic form H1 (V ) is isometric to (dim V )H1 over F , the ideal consisting
of the hyperbolic forms over F in W c (F ) is the principal ideal H1 by Lemma 2.1 (3).
The quotient W (F ) := W c (F )/(H1 ) is called the Witt ring of non-degenerate symmetric
bilinear forms over F . Elements in W (F ) are called Witt classes. Abusing notation, we
shall also write b ∈ W (F ) for the Witt class of b and often call it just the class of b. The
operations in W (F ) (and Wc (F )) shall be denoted by + and ·.
By 1.17, we have
ha, −ai = 0 in W (F )
×
for all a ∈ F and in all characteristics. In particular, h−1i = −h1i = −1 in W (F ), hence
the additive inverse of the Witt class of any non-degenerate symmetric bilinear form b
in W (F ) is represented by the form −b. It follows that if α ∈ W (F ) then there exists a
non-degenerate bilinear form b such that α = b in W (F ).
2. THE WITT AND WITT-GROTHENDIECK RINGS OF SYMMETRIC BILINEAR FORMS 13
Exercise 2.3. (Cf. Scharlau [125], p.22.) Let b be a non-degenerate symmetric

bilinear form on V . Suppose that V = W1 ⊕ W2 with W1 = W1⊥ . Show that
b ⊥ −b ' H(W1 ) ⊥ −b.
In particular, b = H(W1 ) in W
c (F ).
Use this to give another proof that b + (−b) = 0 in W (F ) for every non-degenerate form
b.
The Witt Cancellation Theorem 1.29 allows us to conclude the following.

Proposition 2.4. Let b1 and b2 be anisotropic symmetric bilinear forms. Then the
following are equivalent:
(1) b1 ' b2 .
c (F ).
(2) b1 = b2 in W
(3) b1 = b2 in W (F ).
Proof. The implications (1) ⇒ (2) ⇒ (3) are easy.
(3) ⇒ (1): By definition of the Witt ring, b1 + nH = b2 + mH in W c (F ) for some
n, m ≥ 0. It follows from the definition of the Grothendieck-Witt ring that
b1 ⊥ nH ⊥ b ' b2 ⊥ mH ⊥ b
for some non-degenerate form b. Thus b1 ⊥ nH ⊥ b ⊥ −b ' b2 ⊥ mH ⊥ b ⊥ −b and
b1 ' b2 by Corollary 1.29. ¤
We also have
Corollary 2.5. b = 0 in W (F ) if and only if b is metabolic.
It follows from Proposition 2.4 that every Witt class in W (F ) contains (up to isom-
etry) a unique anisotropic form. As every anisotropic bilinear form is diagonalizable by
Corollary 1.20, we have a ring epimorphism
X X
(2.6) Z[F × /F ×2 ] → W (F ) given by ni (ai F ×2 ) 7→ ni hai i.
i i
Proposition 2.7. A homomorphism of fields F → K induces ring homomorphisms

c (F ) → W
rK/F : W c (K) and rK/F : W (F ) → W (K).
If K/F is purely transcendental then these maps are injective.
Proof. Let b be symmetric bilinear form over F . Define rK/F (b) on K ⊗F Vb by
rK/F (b)(x ⊗ v, y ⊗ w) = xyb(v, w)
for all x, y ∈ K and for all v, w ∈ Vb . This construction is compatible with orthogonal
sums and tensor products of symmetric bilinear forms.
As rK/F (b) is hyperbolic if b is, it follows that b 7→ rK/F (b) induces the desired maps.
These are ring homomorphisms.
The last statement follows by Lemma 1.22. ¤
The ring homomorphisms defined above are called restriction maps. Of course, if
K/F is a field extension then the maps rK/F are the unique homomorphisms such that
rK/F (b) = bK .
3. Chain equivalence
Two non-degenerate diagonal symmetric bilinear forms a = ha1 , a2 , . . . , an i and b =

hb1 , b2 , . . . , bn i, are called simply chain equivalent if either n = 1 and a1 F ×2 = b1 F ×2 or
n ≥ 2 and hai , aj i ' hbi , bj i for some indices i 6= j and ak = bk for every k 6= i, j. Two
non-degenerate diagonal forms a and b are called chain equivalent (we write a ≈ b) if
there is a chain of forms b1 = a, b2 , . . . , bm = b such that bi and bi+1 are simply chain
equivalent for all i = 1, . . . , m − 1. Clearly a ≈ b implies a ' b.
Note as the symmetric group Sn is generated by transpositions, we have ha1 , a2 , . . . , an i ≈
haσ(1) , aσ(2) , . . . , aσ(n) i for every σ ∈ Sn .
Lemma 3.1. Every non-degenerate diagonal form is chain equivalent to an orthogonal
sum of an anisotropic diagonal form and metabolic binary diagonal forms ha, −ai, a ∈ F × .
Proof. By induction, it is sufficient to prove that any isotropic diagonal form b is
chain equivalent to ha, −ai ⊥ b0 for some diagonal form b0 and a ∈ F × . Let {v1 , . . . , vn }
be the orthogonal basis of b and set b(vi , vi ) = ai . Choose an isotropic vector v with the
smallest number k of nonzero P coordinates. Changing the order of the vi if necessary, we
may assume that v = ki=1 ci vi for nonzero ci ∈ F and k ≥ 2. We prove the statement
by induction on k. If k = 2, the restriction of b to the plane F v1 ⊕ F v2 is metabolic
and therefore is isomorphic to ha, −ai for some a ∈ F × by Example 1.23(3), hence b ≈
ha, −ai ⊥ ha3 , . . . , an i.
If k > 2 the vector v10 = c1 v1 + c2 v2 is anisotropic. Complete v10 to an orthogonal
basis {v10 , v20 } of F v1 ⊕ F v2 by Corollary 1.8 and set a0i = b(vi0 , vi0 ), i = 1, 2. Then
ha1 , a2 i ' ha01 , a02 i and b ≈ ha01 , a02 , a3 , . . . , an i. The vector v has k − 1 nonzero coordinates
in the orthogonal basis {v10 , v20 , v3 , . . . , vn }. Applying the induction hypothesis to the
diagonal form ha01 , a02 , a3 , . . . , an i completes the proof. ¤
Lemma 3.2. (Witt Chain Equivalence) Two anisotropic diagonal forms are chain
equivalent if and only if they are isometric.
Proof. Let {v1 , . . . , vn } and {u1 , . . . , un } be two orthogonal bases of the bilinear form
b with b(vi , vi ) = ai and b(ui , ui ) = bi . We must show that ha1 , . . . , an i ≈ hb1 , . . . , bn i.
We do this by double induction on n and the number k of nonzero coefficients of P u1 in the
basis {vi }. Changing the order of the vi if necessary, we may assume that u1 = ki=1 ci vi
for some nonzero ci ∈ F .
If k = 1, i.e., u1 = c1 v1 , the two (n−1)-dimensional subspaces generated by the vi ’s and
ui ’s respectively with i ≥ 2 coincide. By the induction hypothesis and Witt Cancellation
1.29, ha2 , . . . , an i ≈ hb2 , . . . , bn i, hence ha1 , a2 , . . . , an i ≈ ha1 , b2 , . . . , bn i ≈ hb1 , b2 , . . . , bn i.
If k ≥ 2 set v10 = c1 v1 + c2 v2 . As b is anisotropic, a01 = b(v10 , v10 ) is nonzero. Choose an
orthogonal basis {v10 , v20 } of F v1 ⊕ F v2 and set a02 = b(v20 , v20 ). We have ha1 , a2 i ' ha01 , a02 i.
The vector u1 has k − 1 nonzero coordinates in the basis {v10 , v20 , v3 , . . . , vn }. By the
induction hypothesis ha1 , a2 , a3 , . . . , an i ≈ ha01 , a02 , a3 , . . . , an i ≈ hb1 , b2 , b3 , . . . , bn i. ¤
4. STRUCTURE OF THE WITT RING 15
Exercise 3.3. Prove that a diagonalizable metabolic form b is isometric to h1, −1i⊗b0
for some diagonalizable bilinear form b0 .
4. Structure of the Witt ring

In this section, we give a presentation of the Witt-Grothendieck and Witt rings. The
classes of even dimensional anisotropic symmetric bilinear forms generate an ideal I(F )
in the Witt ring. We also derive a presentation for it and its square, I(F )2 .
4.A. The presentation of W c (F ) and W (F ). We turn to determining presentations

c (F ) and W (F ). The generators will be the isometry classes of non-degenerate 1-
of W
dimensional symmetric bilinear forms. The defining relations arise from the following:
¡ ¢
Lemma 4.1. Let a, b ∈ F × and z ∈ D ha, bi . Then ha, bi ' hz, abzi. In particular, if
a + b 6= 0 then
(4.2) ha, bi ' ha + b, ab(a + b)i.
Proof. By Corollary 1.8, we have ha, bi ' hz, di for some d ∈ F × . Comparing
determinants, we must have abF ×2 = dzF ×2 so dF ×2 = abzF ×2 . ¤
The isometry (4.2) is often called the Witt relation.
Define an abelian group W 0 (F ) by generators and relations. Generators are isometry
classes of non-degenerate 1-dimensional symmetric bilinear forms. For any a ∈ F × we
write [a] for the generator — the isometry class of the form hai. Note that [ax2 ] = [a] for
every a, x ∈ F × . The relations are:
(4.3) [a] + [b] = [a + b] + [ab(a + b)]
for all a, b ∈ F × such that a + b 6= 0.
Lemma 4.4. If ha, bi ' hc, di then [a] + [b] = [c] + [d] in W 0 (F ).
×2
Proof. As ha, bi ' hc, di, we have
¡ abF¢ = detha, bi = dethc, di = cdF ×2 and d =
abcz for some z ∈ F . Since c ∈ D ha, bi , there exist x, y ∈ F satisfying c = ax2 + by 2 .
2 ×
If x = 0 or y = 0, the statement is obvious, so we may assume that x, y ∈ F × . It follows

from (4.3) that
[a] + [b] = [ax2 ] + [by 2 ] = [c] + [ax2 by 2 c] = [c] + [d]. ¤
Lemma 4.5. We have [a] + [−a] = [b] + [−b] in W 0 (F ) for all a, b ∈ F × .
Proof. We may assume that a + b 6= 0. From (4.3), we have
[−a] + [a + b] = [b] + [−ab(a + b)], [−b] + [a + b] = [a] + [−ab(a + b)].
The result follows. ¤
If char F 6= 2, the forms ha, −ai and hb, −bi are isometric by Remark 1.16 (ii). There-
fore, in this case Lemma 4.5 follows from Lemma 4.4.
Lemma 4.6. If ha1 , . . . , an i ≈ hb1 , . . . , bn i then [a1 ] + · · · + [an ] = [b1 ] + · · · + [bn ] in
W 0 (F ).
Proof. We may assume that the forms are simply chain equivalent. In this case the
statement follows from Lemma 4.4. ¤
Theorem 4.7. The Grothendieck-Witt group W c (F ) is generated by the isometry
classes of 1-dimensional symmetric bilinear forms that are subject to the defining rela-
tions hai + hbi = ha + bi + hab(a + b)i for all a, b ∈ F × such that a + b 6= 0.
Proof. It suffices to prove that the homomorphism W 0 (F ) → W c (F ) taking [a] to

hai is an isomorphism. As b ⊥ h1i is diagonalizable for any non-degenerate symmetric
bilinear form b by Corollary 1.19, the map is surjective. An element in the kernel is given
by the difference of two diagonal forms b = ha1 , . . . , an i and b0 = ha01 , . . . , a0n i such that
b = b0 in Wc (F ). By the definition of W
c (F ) and Corollary 1.19, there is a diagonal form
b such that b ⊥ b ' b ⊥ b . Replacing b and b0 by b ⊥ b00 and b0 ⊥ b00 respectively, we
00 00 0 00
may assume that b ' b0 . It follows from Lemma 3.1 that b ≈ b1 ⊥ b2 and b0 ≈ b01 ⊥ b02 ,
where b1 , b01 are anisotropic diagonal forms and b2 , b02 are orthogonal sums of metabolic
planes ha, −ai for various a ∈ F × . It follows from the Corollary 1.29 that b1 ' b01 and
therefore b1 ≈ b01 by Lemma 3.2. Note that the dimension of b2 and b02 are equal. By
Lemmas 4.5 and 4.6, we conclude that [a1 ] + · · · + [an ] = [a01 ] + · · · + [a0n ] in W 0 (F ). ¤
Since the Witt class in W (F ) of the hyperbolic plane H1 is equal to h1, −1i by Remark
1.16(iv), Theorem 4.7 yields
Theorem 4.8. The Witt group W (F ) is generated by the isometry classes of 1-
dimensional symmetric bilinear forms that are subject to the following defining relations:
(1) h1i + h−1i = 0.
(2) hai + hbi = ha + bi + hab(a + b)i for all a, b ∈ F × such that a + b 6= 0.
If char F 6= 2, the above is the well-known presentation of the Witt-Grothendieck and
Witt groups first demonstrated by Witt in [146].
4.B. The presentation of I(F ). The Witt-Grothendieck and Witt rings has a nat-
ural filtration that we now describe. Define the dimension map
c (F ) → Z
dim : W given by dim x = dim b1 − dim b2 if x = b1 − b2 .
This is a well-defined map (cf. Equation 2.2).
b ) denote the kernel of this map. As
We let I(F
¡ ¢ ¡ ¢
c (F )
hai − hbi = h1i − hbi − h1i − hai in W
b ) as an abelian group.
for all a, b ∈ F × , the elements h1i − hai with a ∈ F × generate I(F
c (F ) is generated by the elements h1i and h1i − hxi with x ∈ F × .
It follows that W
Let I(F ) denote the image of I(Fb ) in W (F ). If a ∈ F × write hhaiib or simply hhaii
b ) ∩ (H1 ) = 0, we have I(F ) '
for the binary symmetric bilinear form h1, −aib . As I(F
¡ ¢
b )/ I(F
I(F b ) ∩ (H1 ) ' I(F
b ). Then the map W c (F ) → W (F ) induces an isomorphism
b ) → I(F )
I(F given by h1i − hxi 7→ hhxii.
In particular, I(F ) is the ideal in W (F ) consisting of the Witt classes of even dimensional
forms. It is called the fundamental ideal of W (F ) and is generated by the ¡ classes
¢ hhaii
×
with a ∈ ¡F . Note b
that if F → K is a homomorphism of fields then rK/F I(F ) ⊂ I(K) b
¢
and rK/F I(F ) ⊂ I(K).
The relations in Theorem 4.8 can be rewritten as
hhaii + hhbii = hha + bii + hhab(a + b)ii
for a, b ∈ F × with a + b 6= 0. We conclude
Corollary 4.9. The group I(F ) is generated by the isometry classes of 2-dimensional
symmetric bilinear forms hhaii with a ∈ F × subject to the defining relations
(1) hh1ii = 0.
(2) hhaii + hhbii = hha + bii + hhab(a + b)ii for all a, b ∈ F × such that a + b 6= 0.
¡ ¢
b ) n , the nth power of I(F
Let Ibn (F ) := I(F b ). Then Ibn (F ) maps isomorphically onto
I n (F ) := I(F )n , the nth power of I(F ) in W (F ). It defines the filtration
W (F ) ⊃ I(F ) ⊃ I 2 (F ) ⊃ · · · ⊃ I n (F ) ⊃ · · ·
in which we shall be interested.
For convenience, we let Ib0 (F ) = W
c (F ) and I 0 (F ) = W (F ).
We denote the tensor product hha1 ii ⊗ hha2 ii ⊗ · · · ⊗ hhan ii by
hha1 , a2 , . . . , an iib or simply by hha1 , a2 , . . . , an ii
and call a form isometric to such a tensor product a bilinear n-fold Pfister form. (We
call any form isometric to h1i a 0-fold Pfister form.) For n ≥ 1, the isometry classes of
bilinear n-fold Pfister forms generate I n (F ) as an abelian group.
We shall be interested in relations between isometry classes of Pfister forms in W (F ).
We begin with a study of 1- and 2-fold Pfister forms.
Example 4.10. We have hhaii + hhbii = hhabii + hha, bii in W (F ). In particular,
hhaii + hhbii ≡ hhabii mod I 2 (F ).
As the hyperbolic plane is two dimensional, the dimension invariant induces a map
e0 : W (F ) → Z/2Z by b 7→ dim b mod 2.
Clearly, this is a homomorphism with kernel the fundamental ideal I(F ), so induces an
isomorphism
(4.11) ē0 : W (F )/I(F ) → Z/2Z.
By Corollary 1.25, we have a map

dim b
e1 : I(F ) → F × /F ×2 by b 7→ (−1) 2 det b.
called the signed determinant.
The map e1 is a homomorphism

¡ ¢ as det(b ⊥ b0 ) = det b · det b0 and surjective as
hhaii 7→ aF ×2 . Clearly, e1 hha, bii = F ×2 so e1 induces an epimorphism
(4.12) ē1 : I(F )/I 2 (F ) → F × /F ×2 .
We have
Proposition 4.13. The kernel of e1 is I 2 (F ) and the map ē1 : I(F )/I 2 (F ) → F × /F ×2
is an isomorphism.
Proof. Let f1 : F × /F ×2 → I(F )/I 2 (F ) be given by aF ×2 7→ hhaii + I 2 (F ). This
is a homomorphism by Example 4.10 inverse to ē1 , since I(F ) is generated by hhaii,
a ∈ F ×. ¤
For fields of characteristic different than two this is Pfister’s characterization (cf. [113,
Satz 13, Kor.]) of I 2 (F ).
4.C. The presentation of I 2 (F ). We turn to I 2 (F ).
×
Lemma
¡ 4.14.
¢ Let a, b ∈ F . Then hha, bii = 0 in W (F ) if and only if either a ∈ F ×2
×
or b ∈ D hhaii . In particular, hha, 1 − aii = 0 in W (F ) for any a 6= 1 in F .
Proof. Suppose that hhaii is anisotropic. Then hha,
¡ bii =
¢ 0 in W
¡ (F )¢if and only if
bhhaii ' hhaii by Proposition 2.4 if and only if b ∈ G hhaii = D hhaii by Example
1.15. ¤
Isometries of bilinear 2-fold Pfister forms are easily established using isometries of
binary forms. For example, we have
Lemma 4.15. Let a, b ∈ F × and x, y ∈ F . Let z = ax2 + by 2 6= 0. Then
®®
(1) hha, bii ' a, b(y 2 − ax2 ) if y 2 − ax2 6= 0.
(2) hha, bii ' hhz, −abii.
(3) hha, bii ' hhz, abzii.
(4) If z is a square in F then hha, bii is metabolic. In particular, if char F 6= 2 then
hha, bii ' 2H1 .
Proof. (1): Let w = y 2 − ax2 . We have
hha, bii ' h1, −a, −b, abi ' h1, −a, −by 2 , abx2 i ' h1, −a, −bw, abwi ' hha, bwii.
(2): We have
hha, bii ' h1, −a, −b, abi ' h1, −ax2 , −by 2 , abi ' h1, −z, −zab, abi ' hhz, −abii.
(3) follows from (1)and (2) while (4) follows from (2) and Remark 1.16 (ii). ¤
Explicit examples of such isometries are:
Example 4.16. Let a, b ∈ F × then
(1) hha, 1ii is metabolic.
(2) hha, −aii is metabolic.
(3) hha, aii ' hha, −1ii.
(4) hha, bii + hha, −bii = hha, −1ii in W (F ).

We turn to a presentation of I 2 (F ) first done for fields of characteristic not two in
[33] and rediscovered by Suslin (cf. [134]). It is different from that for I(F ) as we need a
new generating relation. Indeed the analogue of the Witt relation will be a consequence
of our new relation and a metabolic relation.
Let I2 (F ) be the abelian group generated by all the isometry classes [b] of bilinear
2-fold Pfister forms b subject to the generating relations:
£ ¤
(1) hh1, 1ii = 0.
£ ¤ £ ¤ £ ¤ £ ¤
(2) hhab, cii + hha, bii = hha, bcii + hhb, cii for all a, b, c ∈ F × .
We call the second relation the cocycle relation.
Remark 4.17. The cocycle relation holds in I 2 (F ): Let a, b, c ∈ F × . Then
hhab, cii + hha, bii = h1, −ab, −c, abci + h1, −a, −b, abi =
h1, 1, −c, abc, −a, −bi = h1, −a, −bc, abci + h1, −b, −c, bci =
hha, bcii + hhb, cii
in I 2 (F ).
We begin by showing that the analogue of the Witt relation is a consequence of the
other two relations.
Lemma 4.18. The relations
£ ¤
(i) hha, 1ii = 0
£ ¤ £ ¤ £ ¤ £ ¤
(ii) hha, cii + hhb, cii = hh(a + b), cii + hh(a + b)ab, cii
holds in I2 (F ) for all a, b, c ∈ F × if a + b 6= 0.
Proof. Applying the cocycle relation to a, a, 1 shows that
£ ¤ £ ¤ £ ¤ £ ¤
hh1, 1ii + hha, aii = hha, aii + hha, 1ii .
The first relation now follows. Applying Lemma 4.15 and the cocycle relation to a, c, c
shows that
(4.19)
£ ¤ £ ¤ £ ¤ £ ¤ £ ¤ £ ¤ £ ¤
hh−a, cii + hha, cii = hhac, cii + hha, cii = hh−a, cii + hha, cii = hh−1, cii
for all c ∈ F × .
Applying the cocycle relation to a(a + b), a, c yields
£ ¤ £ ¤ £ ¤ £ ¤
(4.20) hha + b, cii + hha(a + b), aii = hha(a + b), acii + hha, cii
and to a(a + b), b, c yields
£ ¤ £ ¤ £ ¤ £ ¤
(4.21) hhab(a + b), cii + hha(a + b), bii = hha(a + b), bcii + hhb, cii .
Adding the equations (4.20) and (4.21) and then using the isometries
hha(a + b), aii ' hha(a + b), −bii and hha(a + b), acii ' hha(a + b), −bcii
derived from Lemma 4.15, followed by using equation (4.19) yields

£ ¤ £ ¤ £ ¤ £ ¤
hha, cii + hhb, cii − hh(a + b), cii − hha + b)ab, cii
£ ¤ £ ¤ £ ¤ £ ¤
= hha(a + b), aii + hha(a + b), bii − hha(a + b), acii − hha(a + b), bcii
£ ¤ £ ¤ £ ¤ £ ¤
= hha(a + b), −bii + hha(a + b), bii − hha(a + b), −bcii − hha(a + b), bcii
£ ¤ £ ¤
= hha(a + b), −1ii − hha(a + b), −1ii = 0. ¤
Theorem 4.22. The ideal I 2 (F ) is generated as an abelian group by the isometry
classes hha, bii of bilinear 2-fold Pfister forms for all a, b ∈ F × subject to the generating
relations
(1) hh1, 1ii = 0.
(2) hhab, cii + hha, bii = hha, bcii + hhb, cii for all a, b, c ∈ F × .
Proof. Clearly, we have well-defined homomorphisms
g : I2 (F ) → I 2 (F ) induced by [b] 7→ b
and £ ¤
j : I2 (F ) → I(F ) induced by hha, bii 7→ hhaii + hhbii − hhabii,
the latter being the composition with the inclusion I 2 (F ) ⊂ I(F ) using Example 4.10.
We show that the map g : I2 (F ) → I 2 (F ) is an isomorphism. Define
£ ¤
γ : F × /F ×2 × F × /F ×2 → I2 (F ) by (aF ×2 , bF ×2 ) 7→ hha, bii .
This is clearly well-defined. For convenience, write (a) for aF ×2 . Using (2), we see that
¡ ¢ ¡ ¢ ¡ ¢ ¡ ¢
γ (b), (c) − γ (ab), (c) + γ (a), (bc) − γ (a), (b)
£ ¤ £ ¤ £ ¤ £ ¤
= hhb, cii − hhab, cii + hha, bcii − hha, bii = 0
£ ¤
so γ is a 2-cocycle. By Lemma 4.18, we have hh1, aii = 0 in I2 (F ), so γ is a normalized
2-cocycle. The map γ defines an extension N = F × /F ×2 × I2 (F ) of I2 (F ) by F × /F ×2
with ¡ ¢ ¡ ¢ ¡ ¢
(a), α + (b), β = (ab), α + β + [hha, bii] .
As γ is symmetric, N is abelian. Let
¡ ¢
h : N → I(F ) be defined by (a), α 7→ hhaii + j(α).
We see that the map h is a homomorphism:
¡ ¢ ¡ ¢
h ((a), α) + ((b), β) = h (ab), α + β + [hha, bii]
¡ ¢
= hhabii + j(α) + j(β) + j [hha, bii] = hhaii + hhbii + j(α) + j(β)
¡ ¢ ¡ ¢
= h (a), α + h (b), β .
Thus we have a commutative diagram
0 −−−→ I2 (F ) −−−→ N −−−→ F × /F ×2 −−−→ 0
  
gy
  
hy f1 y
0 −−−→ I 2 (F ) −−−→ I(F ) −−−→ I(F )/I 2 (F ) −−−→ 0

5. THE STIEFEL-WHITNEY MAP 21
where f1 is the isomorphism inverse of ē1 in Proposition 4.13.

Let ¡ ¢
k : I(F ) → N be induced by hhaii 7→ (a), 0 .
Using Lemma 4.15 and Corollary 4.9, we see that k is well-defined as
¡ ¢ ¡ ¢ ¡ ¢ ¡ ¢
(a), 0 + (b), 0 = (ab), [hha, bii] = (ab), [hha + b, ab(a + b)ii]
¡ ¢ ¡ ¢
= (a + b), 0 + (ab(a + b), 0
if a + b 6= 0. As
¡ ¢ ¡ ¢ ¡ ¢ ¡ ¢ ¡ ¢
k hha, bii = k hhaii + hhbii − hhabii = (a), 0 + (b), 0 − (ab), 0
¡ ¢ ¡ ¢ ¡ ¢ ¡ ¢ ¡ ¢ ¡ ¢
= (ab), [hha, bii] − (ab), 0 = (ab), 0 + 1, [hh, a, bii] − (ab), 0 = 1, [hh, a, bii] ,
we have ¡ ¢ ¡ ¢ ¡ ¢
(k ◦ h) (c), [hha, bii] = k hhcii + hha, bii = (c), [hha, bii]) .
¡ ¢
Hence k ◦ h is the identity on N . As (h ◦ k) hhaii = hhaii, the composition h ◦ k is the
identity on I(F ). Thus h is an isomorphism hence so is g. ¤
5. The Stiefel-Whitney map
In this section, we investigate Stiefel-Whitney maps. In the case of fields of charac-
teristic different from two, this was first defined by Milnor. ` We shall use facts about
Milnor K-theory. ` (Cf. §100.) We write k∗ (F ) := n≥0 n (F ) for the graded ring
k
K∗ (F )/2K∗ (F ) := n≥0 Kn (F )/2Kn (F ). Abusing notation, if {a1 , . . . an } is a sym-
bol in Kn (F ), we shall also write it for its coset {a1 , . . . , an } + 2Kn (F ).
`
The associated graded ring GW∗ (F ) = n≥0 I n (F )/I n+1 (F ) of W (F ) with respect to
the fundamental ideal I(F ) is called the graded Witt ring of symmetric bilinear forms.
Note that since 2 · I n (F ) = h1, 1i · I n (F ) ⊂ I n+1 (F ) we have 2 · GW∗ (F ) = 0.
By Example 4.10, the map F × → I(F )/I 2 (F ) defined by a 7→ hhaii + I 2 (F ) is a
homomorphism. By the definition of the Milnor ring and Lemma 4.14, this map gives rise
to a graded ring homomorphism
(5.1) f∗ : k∗ (F ) → GW∗ (F )
taking the symbol {a1 , a2 , . . . , an } to hha1 , a2 , . . . , an ii + I n+1 (F ). Since the graded ring
GW∗ (F ) is generated by the degree one component I(F )/I 2 (F ), the map f∗ is surjective.
Note that the map f0 : k0 (F ) → W (F )/I(F ) is the inverse of the map ē0 and the map
f1 : k1 (F ) → I(F )/I 2 (F ) is the inverse of the map ē1 (cf. Proposition 4.13).
Lemma 5.2. Let hha, bii and hhc, dii be isometric bilinear 2-fold Pfister forms. Then
{a, b} = {c, d} in k2 (F ).
¡ ¢
Proof. If the form hha, bii is metabolic then b ∈ D hhaii or a ∈ F ×2 by Lemma
4.14. In particular, if hha, bii is metabolic then {a, b} = 0 in k2 (F ). Therefore, we
may assume that hha, bii is anisotropic. Using Witt Cancellation 1.29, we see that c =
ax2 + by 2 − abz 2 for some x, y, z ∈ F . If c ∈/ aF ×2 let w = y 2 − az 2 6= 0. Then
hha, bii ' hha, bwii ' hhc, −abwii by Lemma 4.15 and {a, b} = {a, bw} = {c, −abw}
in k2 (F ) by Lemma 100.3. Hence we may assume that a = c. By Witt Cancellation,
¡ ¢
h−b, abi ' h−d, adi so bd ∈ D hhaii , i.e., bd = x2 − ay 2 in F for some x, y ∈ F . Thus
{a, b} = {a, d} by Lemma 100.3. ¤
Proposition 5.3. The homomorphism
e2 : I 2 (F ) → k2 (F ) given by hha, bii 7→ {a, b}
is a well-defined surjection with ker(e2 ) = I 3 (F ). Moreover, e2 induces an isomorphism
ē2 : I 2 (F )/I 3 (F ) → k2 (F ).
Proof. By Lemma 5.2 and the presentation of I 2 (F ) in Theorem 4.22, the map is
well-defined. Since
hha, b, cii = hha, cii + hhb, cii − hhab, cii,
we have I 3 (F ) ⊂ ker(e2 ). As ē2 and f2 are inverses of each other, the result follows. ¤
Let F(F ) be the free abelian group on the set of isometry classes of non-degenerate
1-dimensional symmetric bilinear bilinear forms. Then we have a group homomorphism
w : F(F ) → k∗ (F )[[t]]× given by hai 7→ 1 + {a}t.
If a, b ∈ F × satisfy a + b 6= 0 then by Lemma 100.3, we have
¡ ¢
w hai + hbi = (1 + {a}t)(1 + {b}t)
= 1 + ({a} + {b})t + {a, b}t2
= 1 + ({ab})t + {a, b}t2
= 1 + {ab(a + b)2 }t + {a + b, ab(a + b)}t2
¡ ¢
= w ha + bi + hab(a + b)i .
In particular, w factors through the relation hai + hbi = ha + bi + hab(a + b)i for all
a, b ∈ F × satisfying a + b 6= 0 hence induces a group homomorphism
(5.4) c (F ) → k∗ (F )[[t]]×
w:W
by Theorem 4.7 called the total Stiefel-Whitney map . If b is a non-degenerate symmetric
c (F ) define the total Stiefel-Whitney class w(b) of b
bilinear form and α is its class in W
to be w(α).
Example 5.5. If b is a metabolic plane then b = hai + h−ai in W c (F ) for some

× c (F ) by Example 1.16(iv)), so
a ∈ F . (Note the hyperbolic plane equals h1i + h−1i in W
×
w(b) = 1 + {−1}t as {a, −a} = 1 in k2 (F ) for any a ∈ F .
¡ ¢ ¡ ¢
Lemma 5.6. Let α = h1i − ha1 i · · · h1i − han i in Wc (F ). Let m = 2n−1 . Then
w(α) = (1 + {a1 , . . . , an , −1, . . . , −1 }tm )−1 .

| {z }
m−n
5. THE STIEFEL-WHITNEY MAP 23
Proof. As
X
α= sε haε11 · · · aεnn i,
ε
P
where the sum runs over all ε = (ε1 , . . . , εn ) ∈ {0, 1}n and sε = (−1) i εi , we have
Y X
w(α) = (1 + εi {ai }t)sε .
ε i
Let
Y
h = h(t1 , . . . , tn ) = (1 + ε1 t1 t · · · + · · · + εn tn t)−sε
ε
¡ ¢
in (Z/2Z)[[t]] [[t1 , . . . , tn ]], the ring of power series over Z/2Z in variables t, t1 , . . . , tn .
Substituting zero for any ti in h, yields one so
h = 1 + t1 · · · tn g(t1 , . . . , tn )tn for some g ∈ (Z/2Z[[t]])[[t1 , . . . , tn ]].
As {a, a} = {a, −1}, we have
w(α)−1 = 1 + {a1 , . . . , an }g({a1 }, . . . , {an })tn = 1 + {a1 , . . . , an }g({−1}, . . . , {−1})tn .
We have, with s a variable,
Y X
1 + g(s, . . . , s)tn = h(s, . . . , s) = (1 + εi st)−sε = (1 + st)m = 1 + sm tm
ε i≥1
P
as εi = 1 in Z/2Z exactly m times, so g(s, . . . , s) = (st)m−n and the result follows. ¤
Let w0 (α) = 1 and

X
w(α) = 1 + wi (α)ti
i≥1
c (F ). The map wi : W
for α ∈ W c (F ) → ki (F ) is called the ith Stiefel-Whitney class . Let
c (F ). As w(α + β) = w(α)w(β), for every n ≥ 0, we have the Whitney formula
α, β ∈ W
X
(5.7) wn (α + β) = wi (α)wj (β).
i+j=n
Remark 5.8. Let K/F be a field extension and α ∈ W c (F ). Then

¡ ¢
resK/F wi (α) = wi (αK ) in ki (F ) for all i.
¡ ¢
Corollary 5.9. Let m = 2n−1 . Then wj Ibn (F ) = 0 for j = 1, . . . , m − 1 and
¡ ¢ ¡ ¢
wm : Ibn (F ) → km (F ) is a group homomorphism mapping h1i − ha1 i · · · h1i − han i to
{a1 , . . . , an , −1, . . . , −1 }.
| {z }
m−n
¡ ¢ ¡ ¢
Proof. Let α = h1i − ha1 i · · · h1i − han i . By Lemma 5.6, we have wi (α) = 0 for
i = 1, . . . m − 1. The result follows from the Whitney formula (5.7). ¤
b ) → I(F ) be the isomorphism sending h1i − hai 7→ hhaii. Let w̃m be the
Let j : I(F
composition
−1
j wm |
Ibn (F )
I n (F ) −−→ Ibn (F ) −−−−−→ km (F ).
Corollary 5.9 shows that w̃i = ei for i = 1, 2. The map w̃m : I n (F ) → km (F ) is a group
homomorphism with I n+1 (F ) ⊂ ker(w̃m ) so induces a homomorphism
w̄m : I n (F )/I n+1 (F ) → km (F ).
We have w̄i = ēi for i = 1, 2. The composition w̄m ◦ fn is multiplication by {−1, . . . , −1}.
| {z }
m−n
In particular, w̄1 and w̄2 are isomorphisms, i.e.,
(5.10) I 2 (F ) = ker(w̃1 ) and I 3 (F ) = ker(w̃2 )
and
(5.11) Ib2 (F ) = ker(w1 |I(F
b )) and Ib3 (F ) = ker(w2 |Ib2 (F ) ).
This gives another proof for Propositions 4.13 and 5.3.

Remark 5.12. Let char F 6= 2 and h2F : k2 (F ) → H 2 (F ) be the norm-residue homo-
morphism defined in §101. If b is a non-degenerate symmetric bilinear form then h2 ◦w2 (b)
is the classical Hasse-Witt invariant of b. (Cf. [90], Definition V.3.17, [125], Definition
2.12.7.) More generally, the Stiefel Whitney classes defined above are compatible with
Stiefel-Whitney classes defined by Delzant wi0 in [29], i.e., hi ◦ wi = wi0 for all i ≥ 0.
Example 5.13. Suppose that K is a real-closed field. (Cf. §95.) Then ki (K) = Z/2Z
for all i ≥ 0 and Wc (K) = Z ⊕ Zξ with ξ = h−1i and ξ 2 = 1. The Stiefel-Whitney
map w : W c (F ) → k∗ (K)[[t]]× is then the map n + mξ 7→ (1 + t)m . In particular, if b
is a non-degenerate form then w(b) determines the signature of b. Hence if b and c
are two non-degenerate symmetric bilinear forms over K, we have b ' c if and only if
dim b = dim c and w(b) = w(c).
¡ ¢
It should be noted that if b = hha1 , . . . , an ii that w(b) is not equal to w(α) = w̃ [b]
¡ ¢ ¡ ¢
where α = h1i − ha1 i · · · h1i − han i in W c (F ) as the following exercise shows.
Exercise 5.14. Let m = 2n−1 . If b is the bilinear n-fold Pfister form hha1 , . . . , an ii
then
w(b) = 1 + ({−1, . . . , −1 } + {a1 , . . . , an , −1, . . . , −1 })tm .
| {z } | {z }
m m−n
The following fundamental theorem was proved by Orlov-Vishik-Voevodsky [110] in

the case that char F 6= 2 and by Kato [76] in the case that char F = 2.
Fact 5.15. The map f∗ : k∗ (F ) → GW∗ (F ) is a ring isomorphism.
For i = 0, 1, 2, we have proven that fi is an isomorphism in (4.11), Proposition 4.13 ,
and Proposition 5.3, respectively.
6. BILINEAR PFISTER FORMS 25
6. Bilinear Pfister forms
The isometry classes of tensor products of non-degenerate binary symmetric bilinear

forms representing one are quite interesting. These forms, called Pfister forms, whose
properties over fields of characteristic different from two were discovered by Pfister in [111]
and were named after him in [35], generate a filtration of the Witt ring by the powers of
its fundamental ideal I(F ). Properties of these forms in the case of characteristic two were
first studied by Baeza in [17]. In this section, we derive the main elementary properties
of these forms.
×
¡ By Example
¢ ¡ 1.15,
¢ a bilinear 1-fold Pfister form b = hhaii, a ∈ F , is round, i.e.,
D hhaii = G hhaii . Because of this, the next proposition shows that there are many
round forms and, in particular, bilinear Pfister forms are round.
Proposition 6.1. Let b be a round bilinear form and let a ∈ F × . Then
(1) The form hhaii ⊗ b is also round.
(2) If hhaii ⊗ b is isotropic then either b is isotropic or a ∈ D(b).
Proof. Set c = hhaii ⊗ b.
(1): Since 1 ∈ D(b), it suffices to prove that D(c) ⊂ G(c). Let c be a nonzero value of c.
e
Write c = x − ay for some x, y ∈ D(b). If y = 0, we have c =¡ x ∈ ¢D(b) =
¡ ¢G(b) ⊂ G(c).
Similarly, y ∈ G(c) if x = 0 hence c = −ay ∈ G(c) as −a ∈ G hhaii ⊂ G c .
Now suppose that x and y are nonzero. Since b is round, x, y ∈ G(b) and therefore
c = b ⊥ (−ab) ' b ⊥ (−ayx−1 )b = hhayx−1 ii ⊗ b.
¡ ¢
By Example 1.15, we know that 1 − ayx−1 ∈ G hhayx−1 ii ⊂ G(c). Since x ∈ G(b) ⊂
G(c), we have c = (1 − ayx−1 )x ∈ G(c).
(2): Suppose that b is anisotropic. Since c = b ⊥ (−ab) is isotropic, there exist x, y ∈ D(b)
with x − ay = 0. Therefore a = xy −1 ∈ D(b) as D(b) is closed under multiplication. ¤
Corollary 6.2. Bilinear Pfister forms are round.
Proof. 0-fold Pfister forms are round. ¤
Corollary 6.3. A bilinear Pfister form is either anisotropic or metabolic.
Proof. Suppose that c is an isotropic bilinear Pfister form. We show that c is meta-
bolic by induction on the dimension of the c. We may assume that c = hhaii ⊗ b for a
Pfister form b. If b is metabolic then so is c. By the induction hypothesis, we may assume
that b is anisotropic. By Proposition 6.1 and Corollary 6.2, a ∈ D(b) = G(b). Therefore
ab ' b hence the form c ' b ⊥ (−ab) ' b ⊥ (−b) is metabolic. ¤
Remark 6.4. Note that the only metabolic 1-fold Pfister form is hh1ii. If char F 6= 2
there is only one metabolic bilinear n-fold Pfister form for all n ≥ 1, viz., the hyperbolic
one. It is universal by Corollary 1.26. If char F = 2 then there may exist many metabolic
n-fold Pfister forms for n ≥ 1 including the hyperbolic one.
Example 6.5. If char F = 2, a bilinear Pfister form hha1 , . . . , an ii is anisotropic if
and only if a1 , . . . , an are 2-independent. Indeed [F 2 (a1 , . . . , an ) : F 2 ] < 2n if and only if
hha1 , . . . , an ii is isotropic.
× n
Corollary
¡ n ¢ 6.6. Let char F 6= 2 and z ∈ F . Then 2 hhzii = 0 in W (F ) if and only
if z ∈ D 2 h1i .
¡ ¢
Proof. If z ∈ D 2n h1i then the Pfister form 2n hhzii is isotropic hence metabolic by
Corollary 6.3.
Conversely, suppose that 2n hhzii is metabolic. Then n n n
¡ n 2 ¢h1i = 2n hzi in W (F ). If 2 h1i
is isotropic, it is universal as char F 6= 2, ¡so z ∈¢ D 2 ¡h1i . ¢If 2 h1i is anisotropic then
2n h1i ' 2n hzi by Proposition 2.4 so z ∈ G 2n h1i = D 2n h1i by Corollary 6.2. ¤
As additional corollaries, we have the following two theorems of Pfister (cf. [112]).
The first generalizes the well-known 2-, 4-, and 8-square theorems arising from quadratic
extensions, quaternion algebras, and Cayley algebras.
¡ ¢
Corollary 6.7. D 2n h1i is a group for every non-negative integer n.
The level of a field F is defined to be
s(F ) := min{n | the element − 1 is a sum of n squares}
or infinity if no such integer exists.
Corollary 6.8. The level s(F ) of a field F , if finite, is a power of two.
Proof. Suppose that s(F ) is finite. Then 2n ≤ s(F ) < 2n+1 for some n. By Propo-
sition 6.1 (2), with b = 2n h1i and a = −1, we have −1 ∈ D(b). Hence s(F ) = 2n . ¤
In [112], Pfister also showed that there exist fields of level 2n for all n ≥ 0. (Cf.
Lemma 31.3 below.)
6.A. Chain p-equivalence of bilinear Pfister forms. Since the isometry classes
of 2-fold Pfister forms are easy to deal with, we use them to study n-fold Pfister forms.
We follow the development in [35] which we extend to all characteristics. The case of
characteristic two was also independently done by Arason and Baeza in [3].
Definition 6.9. Let a1 , . . . , an , b1 , . . . , bn ∈ F × with n ≥ 1. We say that hha1 , . . . , an ii
and hhb1 , . . . , bn ii are simply p-equivalent if n = 1 and a1 F ×2 = b1 F ×2 or n ≥ 2 and there
exist i, j = 1, . . . , n such that
hhai , aj ii ' hhbi , bj ii with i 6= j and al = bl for all l 6= i, j.
We say bilinear n-fold Pfister forms b, c are chain p-equivalent if there exist bilinear n-
fold Pfister forms b0 , . . . , bm for some m such that b = b0 , c = bm and bi is simply
p-equivalent to bi+1 for each i = 0, . . . , m − 1.
Chain p-equivalence is clearly an equivalence relation on the set of anisotropic bilinear
forms of the type hha1 , . . . , an ii with a1 , . . . , an ∈ F × and is denoted by ≈. As transposi-
tions generate the symmetric group, we have hha1 , . . . , an ii ≈ hhaσ(1) , . . . , aσ(n) ii for every
permutation σ of {1, . . . , n}. We shall show the following result:
Theorem 6.10. Let hha1 , . . . , an ii and hhb1 , . . . , bn ii be anisotropic. Then
hha1 , . . . , an ii ' hhb1 , . . . , bn ii
if and only if
hha1 , . . . , an ii ≈ hhb1 , . . . , bn ii.
Of course, we need only show isometric anisotropic bilinear Pfister forms are chain
p-equivalent. We shall do this in a number of steps. If b is an n-fold Pfister form then we
can write b ' b0 ⊥ h1i. If b is anisotropic then b0 is unique up to isometry and we call it
the pure subform of b.
Lemma 6.11. Suppose that b = hha1 , . . . , an ii is anisotropic. Let −b ∈ D(b0 ). Then
there exist b2 , . . . , bn ∈ F × such that b ≈ hhb, b2 , . . . , bn ii.
Proof. We induct on n, the case n = 1 being trivial. Let c = hha1 , . . . , an−1 ii so
b ' c0 ⊥ −an c by Witt Cancellation 1.29. Write
0
−b = −x + an y e 0 ),
with − x ∈ D(c e
−y ∈ D(c).
If y = 0 then x 6= 0 and we finish by induction, so we may assume that 0 6= y = y1 + z 2
e 0 ) and z ∈ F . If y1 6= 0 then c ≈ hhy1 , . . . , yn−1 ii for some yi ∈ F × and,
with −y1 ∈ D(c
using Lemma 4.15,
(6.12) c ≈ hhy1 , . . . , yn−1 , an ii ≈ hhy1 , . . . , yn−1 , −an yii ≈ hha1 , . . . , an−1 , −an yii.
This is also true if y1 = 0. If x = 0, we are done. If not c ≈ hhx, x2 , . . . , xn−1 ii some
xi ∈ F × and
b ≈ hhx, x2 , . . . , xn−1 , −an yii ≈ hhan xy, x2 , . . . , xn−1 , −an y + xii
≈ hhan xy, x2 , . . . , xn−1 , bii
by Lemma 4.15(2) as needed. ¤
The argument to establish equation (6.12) yields
Corollary 6.13. Let b = hhx1 , . . . , xn ii and y ∈ D(b). Let z ∈ F × . If b ⊗ hhzii is
anisotropic then hhx1 , . . . , xn , zii ≈ hhx1 , . . . , xn , yzii.
We also have the following generalization of Lemma 4.14:
Corollary 6.14. Let b be an anisotropic bilinear Pfister form over F and let a ∈ F × .
Then ¡hhaii ·¢ b = 0 in W (F ) if and only if either a ∈ F ×2 or b ' hhbii ⊗ c for some
b ∈ D hhaii and bilinear Pfister form c. In the latter case, hha, bii is metabolic.
¡ ¢
Proof. Clearly hha, bii = 0 in W (F ) if b ∈ D hhaii . Conversely, suppose that
hhaii ⊗ b = 0. Hence a ∈ G(b) = D(b) by Corollary 6.2. Write¡ a =¢ x2 − b for some x ∈ F
e 0 ). If b = 0 then a ∈ F ×2 . Otherwise, b ∈ D hhaii and b ' hhbii ⊗ c for
and −b ∈ D(b
some bilinear Pfister form c by Lemma 6.11. ¤
The following generalization of Lemma 6.11 is very useful in computation and is the
key to proving further relations among Pfister forms.
Proposition 6.15. Let b = hha1 , . . . , am ii and c = hhb1 , . . . , bn ii be such that b ⊗ c is
anisotropic. Let −c ∈ D(b ⊗ c0 ) then
hha1 , . . . , am , b1 , . . . , bn ii ≈ hha1 , . . . , am , c1 , c2 , . . . , cn−1 , cii
for some c1 , . . . , cn−1 ∈ F × .
Proof. We induct on n. If n = 1 then −c = yb1 for some −y ∈ D(b) and this

case follows by Corollary 6.13, so assume that n > 1. Let d = hhb1 , . . . , bn−1 ii. Then
e ⊗ c) and
c0 ' bn d ⊥ d0 so bc0 ' bn b ⊗ d ⊥ b ⊗ d0 . Write 0 6= −c = bn y − z with −y ∈ D(b
e ⊗ c0 ). If z = 0 then x 6= 0 and
−z ∈ D(b
hha1 , . . . , am , b1 , . . . , bn ii ≈ hha1 , . . . , am , b1 , . . . , bn−1 , −ybn ii
by Corollary 6.13 and we are done. So we may assume that z 6= 0. By induction
hha1 , . . . , am , b1 , . . . , bn−1 ii ≈ hha1 , . . . , am , c1 , c2 , . . . , cn−2 , zii for some c1 , . . . , cn−2 ∈ F × .
If y = 0, tensoring this by h1, −bn i completes the proof, so we may assume that y 6= 0.
Then
hha1 , . . . , am , b1 , . . . , bn ii ≈ hha1 , . . . , am , b1 , . . . , bn−1 , −ybn ii ≈
hha1 , . . . , am , c1 , . . . , cn−2 , z, −ybn ii ≈ hha1 , . . . , am , c1 , . . . , cn−2 , z − ybn , zybn ii ≈
hha1 , . . . , am , c1 , . . . , cn−2 , c, zybn ii
by Lemma 4.15(2). This completes the proof. ¤
Corollary 6.16. (Common Slot Property) Let hha1 , . . . , an−1 , xii and hhb1 , . . . , bn−1 , yii
be isometric anisotropic bilinear forms. Then there exists a z ∈ F × satisfying
hha1 , . . . , an−1 , zii = hha1 , . . . , an−1 , xii and hhb1 , . . . , bn−1 , zii = hhb1 , . . . , bn−1 , yii.
Proof. Let b = hha1 , . . . , an−1 ii and c = hhb1 , . . . , bn−1 ii. As xb − yc = b0 − c0 in
W (F ), the form xb ⊥ −yc is isotropic. Hence there exists a z ∈ D(xb) ∩ D(yc). The
result follows by Proposition 6.15. ¤
A non-degenerate symmetric bilinear form b is called a general bilinear n-fold Pfister

form if b ' ac for some a ∈ F × and bilinear n-fold Pfister form c. As Pfister forms are
round, a general Pfister form is a Pfister form if and only if it represents one.
Corollary 6.17. Let c and b be general anisotropic bilinear Pfister forms. If c is a
subform of b then b ' c ⊗ d for some bilinear Pfister form d.
Proof. If c = cc1 for some Pfister form c1 and c ∈ F × then c1 is a subform of cb. In
particular, cb represents one so is a Pfister form. Replacing b by cb and c by cc, we may
assume both are Pfister forms.
Let c ' hha1 , . . . , an ii with ai ∈ F × . By Witt Cancellation 1.29, we have c0 is isometric
to a subform of b0 hence b ' hha1 ii ⊗ d1 for some Pfister form d1 by Lemma 6.11. By
induction, there exists a Pfister form dk satisfying b ' hha1 , . . . , ak ii ⊗ dk . By Witt
Cancellation 1.29,¡we have hha1 , . . . , a¢k ii⊗hhak+1 , . . . , an ii0 is a subform of hha1 , . . . , ak ii⊗
d0k so −ak+1 ∈ D hha1 , . . . , ak ii ⊗ d0k . By Proposition 6.15, we complete the induction
step. ¤
Let b and c be general Pfister forms. We say that c divides b if b ' c ⊗ d for some
Pfister form d. The corollary says that c divides b if and only if it is isometric to a subform
of b.
We now proof Theorem 6.10.
Proof. Let a = hha1 , . . . , an ii and b = hhb1 , . . . , bn ii be isometric over F . Clearly

we may assume that n > 1. By Lemma 6.11, we have a ≈ hhb1 , a02 , . . . , a0n ii for some
a0i ∈ F × . Suppose that we have shown a ≈ hhb1 , . . . , bm , a0m+1 , . . . , a0n ii for some m. By
Witt Cancellation 1.29,
hhb1 , . . . , bm ii ⊗ hhbm+1 , . . . , bn ii0 ' hhb1 , . . . , bm ii ⊗ hha0m+1 , . . . , a0n ii0 ,
¡ ¢
so −bm+1 ∈ D hhb1 , . . . , bm ii ⊗ hha0m+1 , . . . , a0n ii0 . By Proposition 6.15, we have
a ≈ hhb1 , . . . , bm+1 , a00m+2 , . . . , a00n ii
for some a00i ∈ F × . This completes the induction step. ¤
We need the following theorem of Arason and Pfister (cf. [11]):

Theorem 6.18. (Hauptsatz) Let 0 6= b be an anisotropic form lying in I n (F ). Then
dim b ≥ 2n .
We shall prove this theorem in Theorem 23.8 below. Using it we show:
Corollary 6.19. Let b and c be two anisotropic general bilinear n-fold Pfister forms.
If b ≡ c mod I n+1 (F ) then b ' ac for some a ∈ F × . In addition, if D(b) ∩ D(c) 6= ∅ then
b ' c.
Proof. Choose a ∈ F × such that b ⊥ −ac is isotropic. By the Hauptsatz, this form
must be metabolic. By Proposition 2.4, we have b ' ac.
Suppose that x ∈ D(b) ∩ D(c). Then b ⊥ −c is isotropic and one can take a = 1. ¤
Theorem 6.20. Let a1 , . . . , an , b1 , . . . , bn ∈ F × . The following are equivalent:
(1) hha1 , . . . , an ii = hhb1 , . . . , bn ii in W (F ).
(2) hha1 , . . . , an ii ≡ hhb1 , . . . , bn ii mod I n+1 (F ).
(3) {a1 , . . . , an } = {b1 , . . . , bn } in Kn (F )/2Kn (F ).
Proof. Let b = hha1 , . . . , an ii and c = hhb1 , . . . , bn ii. As metabolic Pfister forms are
trivial in W (F ) and any bilinear n-fold Pfister form lying in I n+1 (F ) must be metabolic
by the Hauptsatz 6.18, we may assume that b and c are both anisotropic.
(2) ⇒ (1) follows from Corollary 6.19.
(1) ⇒ (3): By Theorem 6.10, we have hha1 , . . . , an ii ≈ hhb1 , . . . , bn ii, so it suffices to show
that (3) holds if
hhai , aj ii ' hhbi , bj ii with i 6= j and al = bl for all l 6= i, j.
As {ai , aj } = {bi , bj } by Proposition 5.3, statement (3) follows.
(3) ⇒ (2) follows from (5.1). ¤
Arason proved and used the Common Slot Property 6.16 in [6] to give an independent
proof in the case of characteristic different from two of Theorem 6.20 which was first
proven in [35].
6.B. Linkage of bilinear Pfister forms. We derive some further properties of

bilinear Pfister forms that we shall need later.
Proposition 6.21. Let b1 and b2 be two anisotropic general bilinear Pfister forms.
Let c be a general r-fold Pfister form with r ≥ 0 that is isometric to a subform of b1
and to a subform of b2 . If i(b1 ⊥ −b2 ) > 2r then there exists a k-fold Pfister form with
k > 0 d such that c ⊗ d is isometric to a subform of b1 and to a subform of b2 . Further
i(b1 ⊥ −b2 ) = 2r+k .
Proof. By Corollary 6.17, there exist Pfister forms d1 and d2 such that b1 ' c ⊗ d1
and b2 ' c ⊗ d2 . Let b = b1 ⊥ −b2 . As b is isotropic, b1 and b2 have a common nonzero
value. Dividing the bi by this nonzero common value, we may assume that the bi are
Pfister forms. We have
b ' c ⊗ (d01 ⊥ −d02 ) ⊥ (c ⊥ −c).
The form c ⊥ −c is metabolic by Example 1.23(2) and i(b) > dim c. Therefore, the form
c ⊗ (d01 ⊥ −d02 ) is isotropic hence there is a ∈ D(c ⊗ d01 ) ∩ D(c ⊗ d02 ). By Proposition 6.15,
we have b1 ' c ⊗ hh−aii ⊗ e1 and b2 ' c ⊗ hh−aii ⊗ e2 for some bilinear Pfister forms e1
and e2 . As
¡ ¢
b ' c ⊗ (e01 ⊥ −e02 ) ⊥ c ⊗ hh−aii ⊥ −c ⊗ hh−aii ,
either i(b) = 2r+1 or we may repeat the argument. The result follows. ¤
If a general bilinear r-fold Pfister form c is isometric to a common subform of two

general Pfister forms b1 and b2 , we call it a linkage of b1 and b2 and say that b1 and b2 are
r-linked. The integer m = max{r | b1 and b2 are r-linked} is called the linkage number
of b1 and b2 . The Proposition says that i(b1 ⊥ −b2 ) = 2m . If b1 and b2 are n-fold Pfister
forms and r = n − 1, we say that b1 and b2 are linked. By Corollary 6.17 the linkage
of any pair of bilinear Pfister forms is a divisor of each. The theory of linkage was first
developed in [35].
If b is a non-degenerate symmetric bilinear form over F then the annihilator of b in
W (F )
annW (F ) (b) := {c ∈ W (F ) | b · c = 0}
is an ideal in W (F ). When b is a Pfister form this ideal has a nice structure that we now
establish. First note that if b is an anisotropic Pfister form and x ∈ D(b) then, as b is
round by Corollary 6.2, we have hhxii ⊗ b ' b ⊥ −xb ' b ⊥ −b is metabolic. It follows
that hhxii ∈ annW (F ) (b). We shall show that these binary forms generate annW (F ) (b)
This will follow from the next result, also known as the Pfister-Witt Theorem.
Proposition 6.22. Let b be an anisotropic bilinear Pfister form and c a non-degenerate
symmetric bilinear form. Then there exists a symmetric bilinear form d satisfying all of
the following:
(1) b · c = b · d in W (F ).
(2) b ⊗ d is anisotropic. Moreover, dim d ≤ dim c and dim d ≡ dim c mod 2.
(3) c − d lies in the subgroup of W (F ) generated by hhxii with x ∈ D(b).
Proof. We prove this by induction on dim c. By the Witt Decomposition Theorem

1.28, we may assume that c is anisotropic. Hence c is diagonalizable by Corollary 1.20, say
c = hx1 , . . . , xn i with xi ∈ F × . If b ⊗ c is anisotropic, the result is trivial, so assume it is
e
isotropic. Therefore, there exist a1 , . . . , an ∈ D(b) not all zero such that a1 x1 +· · ·+an xn =
0. Let bi = ai if ai 6= 0 and bi = 1 otherwise. In particular, bi ∈ G(b) for all i. Let
e = hb1 x1 , . . . , bn xn i. Then c − e = x1 hhb1 ii + · · · + xn hhbn ii with each bi ∈ D(b) as b
is round by Corollary 6.2. Since e is isotropic, we have b · c = b · (e)an in W (F ). As
dim(e)an < dim c, by the induction hypothesis there exists d such that b ⊗ d is anisotropic
and e − d and therefore c − d lies in the subgroup of W (F ) generated by hhxii with
x ∈ D(b). As b ⊗ d is anisotropic, it follows by (1) that dim d ≤ dim c. It follows from
(3) that the dimension of c − d is even. ¤
Corollary 6.23. Let b be an anisotropic bilinear Pfister form. Then annW (F ) (b) is
generated by hhxii with x ∈ D(b).
If b is 2-dimensional, we obtain stronger results first established in [41].
Lemma 6.24. Let b be a binary anisotropic bilinear form over F and c an anisotropic
bilinear form over F such that b ⊗ c is isotropic. Then c ' d ⊥ e for some symmetric
binary bilinear form d annihilated by b and symmetric bilinear form e over F .
Proof. Let {e, f } be a basis for Vb . By assumption there exist vectors v, w ∈ Vc such
that e ⊗ v + f ⊗ w is an isotropic vector for b ⊗ c. Choose a two-dimensional subspace
W ⊂ Vc containing v and w. Since c is anisotropic, so is c|W . In particular, c|W is non-
degenerate hence c = c|W ⊥ c|W ⊥ by Proposition 1.7. As b ⊗ (c|W ) is an isotropic general
2-fold Pfister form it is metabolic by Corollary 6.3. ¤
Proposition 6.25. Let b be a binary anisotropic bilinear form over F and c an
anisotropic form over F . Then there exist symmetric bilinear forms c1 and c2 over F
such that c ' c1 ⊥ c2 with b ⊗ c2 anisotropic and c1 ' d1 ⊥ · · · ⊥ dn where each di is a
binary bilinear form annihilated by b. In particular, if det di = di F ×2 then −di ∈ D(b)
for each i.
Proof. The first statement of the proposition follows from the lemma and the second
from its proof. ¤
Corollary 6.26. Let b be a binary anisotropic bilinear form over F and c an anisotropic
form over F annihilated by b. Then c ' d1 ⊥ · · · ⊥ dn for some symmetric binary forms
di annihilated by b for i = 1, . . . , n.
CHAPTER II
Quadratic Forms
7. Foundations
In this section, we introduce the basic properties of quadratic forms over an arbitrary
field F . Their study arose from the investigation of homogeneous polynomials of degree
two. If the characteristic of F is different from two, then this study and that of bilinear
forms are essentially the same as the diagonal of a bilinear form is a quadratic form and
each determines the other by the polar identity. However, they are different when the
characteristic of F is two. In general, quadratic forms unlike bilinear forms have a rich
geometric flavor. When studying symmetric bilinear forms, we saw that one could easily
reduce to the study of non-degenerate forms. For quadratic forms, the situation is more
complex. The polar form of a quadratic form no longer determines the quadratic form
when the underlying field is of characteristic two. However, the radical of the polar form
is invariant under field extension. This leads to two types of quadratic forms. One is the
study of totally singular quadratic forms, i.e., those whose polar form have radical the
whole underlying space. Such forms need not be trivial in the case of characteristic two.
The other extreme is when the radical of the polar form is as small as possible (which
means of dimension zero or one), this gives rise to non-degenerate quadratic forms. As
in the study of bilinear forms, certain properties are not invariant under base extension.
The most important of these is anisotropy. Analogous to the bilinear case, an anisotropic
quadratic form is one having no nontrivial zero, i.e., no isotropic vectors. Every vector
that is isotropic for the quadratic form is isotropic for its polar form. If the characteristic
is two, the converse is false as every vector is an isotropic vector of the polar form. As in
the previous chapter, we shall base this study on a coordinate free approach and strive to
give uniform proofs in a characteristic free fashion.
Definition 7.1. Let V be a finite dimensional vector space over F . A quadratic form
on V is a map ϕ : V → F satisfying
(1) ϕ(av) = a2 ϕ(v) for all v ∈ V and a ∈ F .
(2) (Polar Identity) bϕ : V × V → F defined by
bϕ (v, w) = ϕ(v + w) − ϕ(v) − ϕ(w)
is a bilinear form.
The bilinear form bϕ is called the polar form of of ϕ. We call dim V the dimension of the
quadratic form and also write it as dim ϕ. We write ϕ is a quadratic form over F if ϕ is
a quadratic form on a finite dimensional vector space over F and denote the underlying
space by Vϕ .
33
34 II. QUADRATIC FORMS
Note that the polar form of a quadratic form is automatically symmetric and even
alternating if char F = 2. If b : V × V → F is a bilinear form (not necessarily symmetric),
let ϕb : V → F be defined by ϕb (v) = b(v, v) for all v ∈ V . Then ϕb is a quadratic form
and its polar form bϕb is b + bt . We call ϕb the associated quadratic form of b.
In particular, if b is symmetric, the composition b 7→ ϕb 7→ bϕb is multiplication by 2
as is the composition ϕ 7→ bϕ 7→ ϕbϕ .
Definition 7.2. Let ϕ and ψ be two quadratic
¡ ¢ forms. An isometry f : ϕ → ψ is a
linear map f : Vϕ → Vψ such that ϕ(v) = ψ f (v) for all v ∈ Vϕ . If such an isometry
exists, we write ϕ ' ψ and say that ϕ and ψ are isometric.
Example 7.3. If ϕ is a quadratic form over F and v ∈ V satisfies ϕ(v) 6= 0 then the
(hyperplane) reflection
τv : ϕ → ϕ given by w 7→ w − bϕ (v, w)ϕ(v)−1 v
is an isometry.
Let V be a finite dimensional vector space over F . Define the hyperbolic form on V
to be H(V ) = ϕH on V ⊕ V ∗ with
ϕH (v, f ) := f (v)
for all v ∈ V and f ∈ V ∗ . Note that the polar form of ϕH is bϕH = H1 (V ). If ϕ is a
quadratic form isometric to H(W ) for some vector space W , we call ϕ a hyperbolic form.
The form H(F ) is called the hyperbolic plane and we denote it simply by H. If ϕ ' H, two
vectors e, f ∈ Vϕ satisfying ϕ(e) = ϕ(f ) = 0 and bϕ (e, f ) = 1 are called a hyperbolic pair.
Let ϕ be a quadratic form on V and {v1 , . . . , vn } be a basis for V . Let aii = ϕ(vi ) for
all i and (
bϕ (vi , vj ) for all i < j
aij =
0 for all i > j.
As n
X X
ϕ( xi v i ) = aij xi xj ,
i=1 i,j
the homogeneous polynomial on the right hand side as well as the matrix (aij ) determined
by ϕ completely determines ϕ.
Notation 7.4. (1) Let a ∈ F . The quadratic form on F given by ϕ(v) = av 2 for all
v ∈ F will be denoted by haiq or simply hai.
(2) Let a, b ∈ F . The two dimensional quadratic form on F 2 given by ϕ(x, y) = ax2 +
xy + by 2 will be denoted by [a, b]. The corresponding matrix for ϕ in the standard basis
is µ ¶
a 1
A= ,
0 b
while the corresponding matrix for bϕ is
µ ¶
2a 1
= A + At .
1 2b
7. FOUNDATIONS 35
Remark 7.5. Let ϕ be a quadratic form on V over F . Then the associated polar form
bϕ is not the zero form if and only if there are two vectors v, w in V satisfying b(v, w) = 1.
In particular, if ϕ is a nonzero binary form then ϕ ' [a, b] for some a, b ∈ F .
Example 7.6. Let {e, f } be a hyperbolic pair for H. Using the basis {e, ae + f }, we
have H ' [0, 0] ' [0, a] for any a ∈ F .
Example 7.7. Let char F = 2 and ℘ : F → F be the Artin-Schreier map ℘(x) =
x2 + x. Let a ∈ F . Then the quadratic form [1, a] is isotropic if and only if a ∈ ℘(F ).
Let V be a finite dimension vector space over F . The set Quad(V) of quadratic forms
on V is a vector space over F . We have linear maps
Bil(V ) → Quad(V ) given by b 7→ ϕb
and
Quad(V ) → Sym(V ) given by ϕ 7→ bϕ .
Restricting the first map to Sym(V ) and composing shows the compositions
Sym(V ) → Quad(V ) → Sym(V ) and Quad(V ) → Sym(V ) → Quad(V )
are multiplication by 2. In particular, if char F 6= 2 the map Quad(V ) → Sym(V ) given
by ϕ 7→ 12 bϕ is an isomorphism inverse to the map Sym(V ) → Quad(V ) by b 7→ ϕb . For
this reason, we shall usually identify quadratic forms and symmetric bilinear forms over
a field of characteristic different from two.
The correspondence between quadratic forms on a vector space V of dimension n and
matrices defines a linear isomorphism Quad(V ) → Tn (F ), where Tn (F ) is the vector space
of n×n-upper triangular matrices. Therefore by the surjectivity of the linear epimorphism
Mn (F ) → Tn (F ) given by (aij ) 7→ (bij ) with bij = aij +aji for all i < j, and bii = aii for all
i, and bij = 0 for all j < i implies that the linear map Bil(V ) → Quad(V ) given by b 7→ ϕb
is also surjective. We, therefore, have an exact sequence
0 → Alt(V ) → Bil(V ) → Quad(V ) → 0.
Exercise 7.8. The natural exact sequence
V
0 → 2 (V ∗ ) → V ∗ ⊗F V ∗ → S 2 (V ∗ ) → 0
can be identified with the sequence above via the isomorphism
S 2 (V ∗ ) → Quad(V ) given by f · g 7→ ϕf ·g : v 7→ f (v)g(v).
If ϕ, ψ ∈ Quad(V), we say ϕ is similar to ψ if there exists an a ∈ F × such that ϕ ' aψ.

Let ϕ be a quadratic form on V . A vector v ∈ V is called anisotropic if ϕ(v) 6= 0 and
isotropic if v 6= 0 and ϕ(v) = 0. We call ϕ anisotropic if there are no isotropic vectors in
V and isotropic if there are.
If W ⊂ V is a subspace, the restriction of ϕ on W is the quadratic form whose polar
form is given by bϕ|W = bϕ |W . It is denoted by ϕ|W and called a subform of ϕ. Define
W ⊥ to be the orthogonal complement of W relative to the polar form of ϕ. The space W
is called totally isotropic if ϕ|W = 0. If this is the case then bϕ |W = 0.
Example 7.9. If F is algebraically closed then any homogeneous polynomial in more

than one variable has a nontrivial zero. In particular, up to isometry, the only anisotropic
quadratic forms over F are 0 and h1i.
Remark 7.10. Let ϕ be a quadratic form on V over F . If ϕ = ϕb for some symmetric
bilinear form b then ϕ is isotropic if and only if b is. In addition, if char F 6= 2 then ϕ is
isotropic if and only if its polar form bϕ is. However, if char F = 2 then every 0 6= v ∈ V
is an isotropic vector for bϕ .
Let ψ be a subform of a quadratic form ϕ. The restriction of ϕ on (Vψ )⊥ is denoted

by ψ ⊥ and is called the complementary form of ψ in ϕ. If Vϕ = W ⊕ U is a direct sum of
vector spaces with W ⊂ U ⊥ , we write ϕ = ϕ|W ⊥ ϕ|U and call it an internal orthogonal
sum. So ϕ(w + u) = ϕ(w) + ϕ(u) for all w ∈ W and u ∈ U . Note that ϕ|U is a subform
of (ϕ|W )⊥ .
Remark 7.11. Let ϕ be a quadratic form with rad bϕ = 0. If ψ is a subform of ϕ
then by Proposition 1.6, we have dim ψ ⊥ = dim ϕ − dim ψ and therefore ψ ⊥⊥ = ψ.
Let ϕ be a quadratic form on V . We say that ϕ is totally singular if its polar form bϕ
is zero. If char F 6= 2 then ϕ is totally singular if and only if ϕ is the zero quadratic form.
If char F = 2 this may not be true. Define the quadratic radical of ϕ by
rad ϕ := {v ∈ rad bϕ | ϕ(v) = 0}.
This is a subspace of rad bϕ . We say that ϕ is regular if rad ϕ = 0. If char F 6= 2 then
rad ϕ = rad bϕ . In particular, ϕ is regular if and only if its polar form is non-degenerate.
If char F = 2, this may not be true.
Example 7.12. Every anisotropic quadratic form is regular.
Clearly, if f : ϕ → ψ is an isometry of quadratic forms then f (rad bϕ ) = rad bψ and
f (rad ϕ) = rad ψ.
Let ϕ be a quadratic form on V and : V → V / rad ϕ the canonical epimorphism.
Let ϕ denote the quadratic form on V given by ϕ(v) := ϕ(v) for all v ∈ V . In particular,
the restriction of ϕ to rad bϕ / rad ϕ determines an anisotropic quadratic form. We have:
Lemma 7.13. Let ϕ be a quadratic form on V and W any subspace of V satisfying
V = rad ϕ ⊕ W . Then
ϕ = ϕ|rad ϕ ⊥ ϕ|W = 0|rad ϕ ⊥ ϕ|W .
with ϕ|W ' ϕ the induced quadratic form on V / rad ϕ. In particular, ϕ|W is unique up to
isometry.
If ϕ is a quadratic form, the form ϕ|W , unique up to isometry will be called its regular
part. The subform ϕ|W in the lemma is regular but bϕ|W may be degenerate if char F = 2.
To obtain a further orthogonal decomposition of a quadratic form, we need to look at the
regular part. The key is the following:
Proposition 7.14. Let ϕ be a regular quadratic form on V . Suppose that V contains
an isotropic vector v. Then there exists a two-dimensional subspace W of V containing v
such that ϕ|W ' H.
7. FOUNDATIONS 37
Proof. As rad ϕ = 0, we have v ∈ / rad bϕ . Thus there exists a vector w ∈ V such that
a = bϕ (v, w) 6= 0. Replacing v by a−1 v, we may assume that a = 1. Let W = F v ⊕ F w.
Then v, w − ϕ(w)v is a hyperbolic pair. ¤
We say that any isotropic regular quadratic form splits off a hyperbolic plane.
If K/F is a field extension let ϕK be the quadratic form on VK defined by ϕK (x⊗v) :=
x2 ϕ(v) for all x ∈ K and v ∈ V with polar form bϕK = (bϕ )K . Although (rad bϕ )K =
rad(bϕ )K , we only have (rad ϕ)K ⊂ rad ϕK with inequality possible.
Remark 7.15. If K/F is a field extension and ϕ a quadratic form over F then ϕ is
regular if ϕK is.
The following is a useful observation. The proof analogous to that for Lemma 1.22
shows:
Lemma 7.16. Let ϕ be an anisotropic quadratic form over F . If K/F is purely tran-
scendental then ϕK is anisotropic.
7.A. Non-degenerate quadratic forms. To define non-degeneracy, we use the fol-

lowing lemma.
Lemma 7.17. Let ϕ be a quadratic form over F . Then the following are equivalent:
(1) ϕK is regular for every field extension K/F .
(2) ϕK is regular over an algebraically closed field K containing F .
(3) ϕ is regular and dim rad bϕ ≤ 1.
Proof. (1) ⇒ (2) is trivial.
(2) ⇒ (3): As (rad ϕ)K ⊂ rad ϕK = 0, we have rad ϕ = 0. To show the second
statement, we may assume that F is algebraically closed. As ϕ|rad bϕ = ϕ|rad bϕ / rad ϕ is
anisotropic and over an algebraically closed field any quadratic form of dimension greater
than one is isotropic, dim rad bϕ ≤ 1.
(3) ⇒ (1): Suppose that rad ϕK 6= 0. Then rad ϕK = rad(bϕ )K = (rad bϕ )K is
one dimensional. Let 0 6= v ∈ rad bϕ . Then v ∈ rad ϕK hence ϕ(v) = 0 contradicting
rad ϕ = 0. ¤
Definition 7.18. A quadratic form ϕ over F is called non-degenerate if the equivalent
conditions of the lemma are satisfied.
Remark 7.19. If K/F is a field extension then ϕ is non-degenerate if and only if ϕK
is non-degenerate by Lemma 7.17.
This definition of a non-degenerate quadratic form agrees with the one given in [87].
It is different than that found in some other texts. The geometric characterization of this
definition of non-degeneracy explains our definition. In fact, if ϕ is a quadratic form on
V of dimension at least two then the following are equivalent:
(1) The quadratic form ϕ is non-degenerate.
(2) The projective quadric Xϕ associated to ϕ is smooth. (Cf. Proposition 22.1.)
(3) The even Clifford algebra C0 (ϕ) (cf. §11 below) of ϕ is separable (i.e., is a product
of finite dimensional simple algebras each central over a separable field extension
of F ). (Cf. Proposition 11.6.)
(4) The group scheme SO(ϕ) of all isometries of ϕ identical on rad ϕ is reductive
(semi-simple if dim ϕ ≥ 3 and simple if dim ϕ ≥ 5). (Cf. [87, Chapter VI].)
Proposition 7.20. (i) The form hai is non-degenerate if and only if a ∈ F × .
(ii) The form [a, b] is non-degenerate if and only if 1 − 4ab 6= 0. In particular this binary
quadratic form as well as its polar form is always non-degenerate if char F = 2.
(iii) Hyperbolic forms are non-degenerate.
(iv) Every binary isotropic non-degenerate quadratic form is isomorphic to H.
Proof. (i) and (iii) are clear.
(ii): This follows by computing the determinant of the matrix representing the polar form
corresponding to [a, b]. (Cf. Notation 7.4.)
(iv) follows by Proposition 7.14. ¤
Let ϕi be a quadratic formòn Vi for i = 1, 2. Then their external orthogonal sum is

defined by ϕ := ϕ1 ⊥ ϕ2 on V1 V2 given by
¡ ¢
ϕ (v1 , v2 ) := ϕ1 (v1 ) + ϕ2 (v2 )
for all vi ∈ Vi , i = 1, 2. Note that bϕ1 ⊥ϕ2 = bϕ1 ⊥ bϕ2 .
Remark 7.21. Let char F 6= 2. Let ϕ and ψ be quadratic forms over F .
(1) The form ϕ is non-degenerate if and only if ϕ is regular.
(2) If ϕ and ψ are both non-degenerate then ϕ ⊥ ψ is non-degenerate as bϕ⊥ψ =
bϕ ⊥ bψ .
Remark 7.22. Let char F = 2. Let ϕ and ψ be quadratic forms over F .
(1) If dim ϕ is even then ϕ is non-degenerate if and only if its polar form bϕ is
non-degenerate.
(2) If dim ϕ is odd then ϕ is non-degenerate if and only if dim rad bϕ = 1 and ϕ|rad bϕ
is nonzero.
(3) If ϕ and ψ are non-degenerate quadratic forms over F at least one of which is of
even dimension then ϕ ⊥ ψ is non-degenerate.
The important analogue of Proposition 1.7 is immediate (using Lemma 7.17 for the
last statement):
Proposition 7.23. Let ϕ be a quadratic form on V . Let W be a vector subspace such
that bϕ|W is a non-degenerate bilinear form. Then ϕ|W is non-degenerate and ϕ = ϕ|W ⊥
ϕ|W ⊥ . In particular, (ϕ|W )⊥ = ϕ|W ⊥ . Further, if ϕ is also non-degenerate then so is
ϕ|W ⊥ .
7. FOUNDATIONS 39
Example 7.24. Suppose that char F = 2 and a, b, c ∈ F . Let ϕ = [c, a] ⊥ [c, b] and
{e, f, e0 , f 0 } be a basis for Vϕ satisfying ϕ(e) = c = ϕ(e0 ), ϕ(f ) = a, ϕ(f 0 ) = b, and
bϕ (e, f ) = 1 = b(e0 , f 0 ). Then in the basis {e, f + f 0 , e + e0 , f 0 }, we have
ϕ ' [c, a] ⊥ [c, b] ' [c, a + b] ⊥ H
by Example 7.6.
If n is a non-negative integer and ϕ is a quadratic form over F , we let
nϕ := ϕ ⊥ · · · ⊥ ϕ .
| {z }
n
In particular, if n is an integer, we do not interpret nϕ with n viewed in the field. For

example, if V is an n-dimensional vector space, H(V ) ' nH.
We denote ha1 iq ⊥ · · · ⊥ han iq by
ha1 , . . . , an iq or simply ha1 , . . . , an i.
So ϕ ' ha1 , . . . , an i if and only if Vϕ has an orthogonal basis. If Vϕ has an orthogonal
basis, we say ϕ is diagonalizable.
Remark 7.25. Suppose that char F = 2 and ϕ is a quadratic form over F . Then ϕ
is diagonalizable if and only if ϕ is totally singular, i.e., its polar form bϕ = 0. If this is
the case then every basis for Vϕ is orthogonal. In particular, there are no diagonalizable
non-degenerate quadratic forms of dimension greater than one.
Exercise 7.26. A quadratic form ϕ is diagonalizable if and only if ϕ = ϕb for some
symmetric bilinear form b.
Example 7.27. Suppose that char F 6= 2. If a ∈ F × then ha, −ai ' H.
Example 7.28. (Cf. Example 1.11.) Let char F = 2 and ϕ = h1, ai with a 6= 0. If
{e, f } is the basis for Vϕ with ϕ(e) = 1 and ϕ(f ) = a then computing on the orthogonal
basis {e, xe + yf } with x, y ∈ F , y 6= 0 shows ϕ ' h1, x2 + ay 2 i. Consequently, h1, ai '
h1, bi if and only if b = x2 + ay 2 with y 6= 0.
7.B. Structure theorems for quadratic forms. We wish to decompose a qua-
dratic form over a field F into an orthogonal sum of nice subforms. We begin with
non-degenerate quadratic forms with large totally isotopic subspaces. Unlike the case of
symmetric bilinear forms in characteristic two, two-dimensional non-degenerate isotropic
quadratic forms are hyperbolic.
Proposition 7.29. Let ϕ be an 2n-dimensional non-degenerate quadratic form on V .
Suppose that V contains a totally isotropic subspace W of dimension n. Then ϕ ' nH.
Conversely, every hyperbolic form of dimension 2n contains a totally isotropic subspace
of dimension n.
Proof. Let 0 6= v ∈ W . Then by Proposition 7.14 there exists a two dimensional
subspace V1 of V containing v with ϕ|V1 a non-degenerate subform isomorphic to H. By
Proposition 7.23, this subform splits off as an orthogonal summand. Since ϕ|V1 is non-
degenerate, W ∩V1 is one dimensional, so dim W ∩V1⊥ = n−1. The first statement follows
by induction applied to the totally isotropic subspace W ∩ V1⊥ of V1⊥ . The converse is
easy. ¤
We turn to splitting off anisotopic subforms of regular quadratic forms. It is convenient
to write these decompositions separately for fields of characteristic two and not two.
Proposition 7.30. Let char F 6= 2 and ϕ a quadratic form on V over F . Then there
exists an orthogonal basis for V . In particular, there exist one dimensional subspaces
Vi ⊂ V , 1 ≤ i ≤ n for some n and an orthogonal decomposition
ϕ = ϕ|rad bϕ ⊥ ϕ|V1 ⊥ · · · ⊥ ϕ|Vn
with ϕ|V1 ' hai i, ai ∈ F × for all i = 1, . . . , n. In particular
ϕ ' rh0i ⊥ ha1 , . . . , an i
with r = dim rad bϕ .
Proof. We may assume that ϕ 6= 0. Hence there exists an anisotropic vector 0 6=
v ∈ V . As bϕ|F v is non-degenerate, ϕ|F v splits off as an orthogonal summand of ϕ by
Proposition 7.23. The result follows easily by induction. ¤
Corollary 7.31. Suppose that char F 6= 2. Then every quadratic form over F is
diagonalizable.
Proposition 7.32. Let char F = 2 and ϕ a quadratic form on V over F . Then there
exists two dimensional subspaces Vi ⊂ V , 1 ≤ i ≤ n for some n, a subspace W ⊂ rad bϕ ,
and an orthogonal decomposition
ϕ = ϕ|rad(ϕ) ⊥ ϕ|W ⊥ ϕ|V1 ⊥ · · · ⊥ ϕ|Vn
with ϕ|Vi ' [ai , bi ] non-degenerate, ai , bi ∈ F for all i = 1, . . . , n. Moreover, ϕ|W is
anisotropic, diagonalizable, and is unique up to isometry. In particular,
ϕ ' rh0i ⊥ hc1 , . . . , cs i ⊥ [a1 , b1 ] ⊥ · · · ⊥ [an , bn ]
with r = dim rad ϕ and s = dim W , ci ∈ F × , 1 ≤ i ≤ s.
Proof. Let W ⊂ V be a subspace such that rad bϕ = rad ϕ ⊕ W and V 0 ⊂ V a
subspace such that V = rad bϕ ⊕ V 0 . Then ϕ = ϕ|rad(ϕ) ⊥ ϕ|W ⊥ ϕ|V 0 . The form ϕ|W
is diagonalizable as bϕ|W = 0 and anisotropic as W ∩ rad ϕ = 0. By Lemma 7.13, the
form ϕ|W = (ϕ|rad bϕ )|W is unique up to isometry. So to finish we need only show that
ϕ|V 0 is an orthogonal sum of non-degenerate binary subforms of the desired isometry
type. We may assume that V 0 6= {0}. Let 0 6= v ∈ V 0 . Then there exists 0 6= v 0 ∈ V 0
such that c = bϕ (v, v 0 ) 6= 0. Replacing v 0 by c−1 v 0 , we may assume that bϕ (v, v 0 ) = 1. In
particular, ϕ|F v⊕F v0 ' [ϕ(v), ϕ(v 0 )]. As [ϕ(v), ϕ(v 0 )] and its polar form are non-degenerate
by Proposition 7.20, the subform ϕ|F v⊕F v0 is an orthogonal direct summand of ϕ by
Proposition 7.23. The decomposition follows by Lemma 7.13 and induction. ¤
Corollary 7.33. Let char F = 2 and ϕ a non-degenerate quadratic form over F .
(1) If dim ϕ = 2n then
ϕ ' [a1 , b1 ] ⊥ · · · ⊥ [an , bn ]
for some ai , bi ∈ F , i = 1, . . . , n.
8. WITT’S THEOREMS 41
(2) If dim ϕ = 2n + 1 then

ϕ ' hci ⊥ [a1 , b1 ] ⊥ · · · ⊥ [an , bn ]
for some ai , bi ∈ F , i = 1, . . . , n, and c ∈ F × unique up to F ×2 .
Example 7.34. Suppose that F is quadratically closed of characteristic two. Then
every anisotropic form is isometric to 0, h1i or [1, a] with a ∈ F \ ℘(F ) where ℘ : F → F
is the Artin-Schreier map.
Exercise 7.35. Every non-degenerate quadratic form over a separably closed field F
is isometric to nH or hai ⊥ nH for some n ≥ 0 and a ∈ F × .
8. Witt’s Theorems
As with the bilinear case, the classical Witt theorems are more delicate to ascertain
over fields of arbitrary characteristic. We shall give characteristic free proofs of these.
The basic Witt theorem is the Witt Extension Theorem (cf. Theorem 8.3 below). Witt’s
original theorem (cf. [146]) has been generalized in various ways. We use one similar to
that given by Kneser (cf. [79, Th. 1.2.2]). We construct the quadratic Witt group of even
dimensional anisotropic quadratic forms and use the Witt theorems to study this group.
To obtain further decompositions of a quadratic form, we need generalizations of the
classical Witt theorems for bilinear forms over fields of characteristic different from two.
Let ϕ be a quadratic form on V . Let v and v 0 in V satisfy ϕ(v) = ϕ(v 0 ). If the vector
v̄ = v − v 0 is anisotropic then the reflection (cf. Example 7.3) τv̄ : ϕ → ϕ satisfies
(8.1) τv̄ (v) = v 0 .
What if v̄ is isotropic?
Lemma 8.2. Let ϕ be a quadratic form on V with polar form b. Let v and v 0 lie in
V and v̄ = v − v 0 . Suppose that ϕ(v) = ϕ(v 0 ) and ϕ(v̄) = 0. If w ∈ V is anisotropic and
satisfies both b(w, v) and b(w, v 0 ) are nonzero then the vector w0 = v −τw (v 0 ) is anisotropic
and (τw ◦ τw0 )(v) = v 0 .
Proof. As w0 = v̄ + b(v 0 , w)ϕ(w)−1 w, we have
¡ ¢
ϕ(w0 ) = ϕ(v̄) + b v̄, b(v 0 , w)ϕ(w)−1 w + b(v 0 , w)2 ϕ(w)−1
= b(v, w)b(v 0 , w)ϕ(w)−1 6= 0.
It follows from (8.1) that τw0 (v) = τw (v 0 ) hence the result. ¤
Theorem 8.3. (Witt Extension Theorem) Let ϕ and ϕ0 be isometric quadratic forms
on V and V 0 respectively. Let W ⊂ V and W 0 ⊂ V 0 be subspaces such that W ∩ rad bϕ = 0
and W 0 ∩ rad bϕ0 = 0. Suppose that there is an isometry α : ϕ|W → ϕ0 |W 0 . Then there
exists an isometry α̃ : ϕ → ϕ0 such that α̃(W ) = W 0 and α̃|W = α.
Proof. It is sufficient to treat the case V = V 0 and ϕ = ϕ0 . Let b denote the polar
form of ϕ. We proceed by induction on n = dim W , the case n = 0 being obvious.
Suppose that n > 0. In particular, ϕ is not identically zero. Let u ∈ V satisfy ϕ(u) 6= 0.
As dim W ∩ (F u)⊥ ≥ n − 1, there exists a subspace W0 ⊂ W of codimension one with
W0 ⊂ (F u)⊥ . Applying the induction hypothesis to β = α|W0 : ϕ|W0 → ϕ|α(W0 ) , there

exists an isometry β̃ : ϕ → ϕ satisfying β̃(W0 ) = α(W0 ) and β̃|W0 = β. Replacing W 0 by
β̃ −1 (W 0 ), we may assume that W0 ⊂ W 0 and α|W0 is the identity.
Let v be any vector in W \ W0 and set v 0 = α(v) ∈ W 0 . It suffices to find an isometry
γ of ϕ such that γ(v) = v 0 and γ|W0 = Id, the identity on W0 . Let v̄ = v − v 0 as above
and S = W0⊥ . Note that for every w ∈ W0 , we have α(w) = w, hence
¡ ¢
b(v̄, w) = b(v, w) − b α(v), α(w) = 0,
i.e., v̄ ∈ S.
Suppose that ϕ(v̄) 6= 0. Then τv̄ (v) = v 0 using (8.1). Moreover, τv̄ (w) = w for every
w ∈ W0 as v̄ is orthogonal to W0 . Then γ = τv̄ works. So we may assume that ϕ(v̄) = 0.
We have
0 = ϕ(v̄) = ϕ(v) − b(v, v 0 ) + ϕ(v 0 ) = b(v, v) − b(v, v 0 ) = b(v, v̄),
i.e., v̄ is orthogonal to v. Similarly, v̄ is orthogonal to v 0 .
By Proposition 1.6, the map lW : V → W ∗ is surjective. In particular, there exists
u ∈ V such that b(u, W0 ) = 0 and b(u, v) = 1. In other words, v is not orthogonal to
S, i.e., the intersection H = (F v)⊥ ∩ S is a subspace of codimension one in S. Similarly,
H 0 = (F v 0 )⊥ ∩ S is also a subspace of codimension one in S. Note that v̄ ∈ H ∩ H 0 .
Suppose that there exists an anisotropic vector w ∈ S such that w ∈ / H 0.
/ H and w ∈
0
By Lemma 8.2, we have (τw ◦ τw0 )(v) = v where
w0 = v − τw (v 0 ) = v̄ + b(v 0 , w)ϕ(w)−1 w ∈ S.
As w, w0 ∈ S, the map τw ◦ τw0 is the identity on W0 . Setting γ = τw ◦ τw0 produces the
desired extension. Consequently, we may assume that ϕ(w) = 0 for every w ∈ S \(H ∪H 0 ).
Case 1: |F | > 2.
Let w1 ∈ H ∩ H 0 and w2 ∈ S \ (H ∪ H 0 ). Then aw1 + w2 ∈ S \ (H ∪ H 0 ) for any a ∈ F ,
so by assumption
0 = ϕ(aw1 + w2 ) = a2 ϕ(w1 ) + ab(w1 , w2 ) + ϕ(w2 ).
Since
¡ |F | > 2, ¢we must have ϕ(w1 ) = b(w1 , w2 ) = ϕ(w2 ) = 0. So ϕ(H ∩ H 0 ) = 0,
ϕ S \ (H ∪ H 0 ) = 0, and H ∩ H 0 is orthogonal to S \ (H ∪ H 0 ), (i.e., b(x, y) = 0 for
all x ∈ H ∩ H 0 and y ∈ S \ (H ∪ H 0 )).
Let w ∈ H and w0 ∈ S \ (H ∪ H 0 ). As |F | > 2, we see that w + aw0 ∈ S \ (H ∪ H 0 ) for
some a ∈ F . Hence the set S \(H ∪H 0 ) generates S. Consequently, H ∩H 0 is orthogonal to
S. In particular, b(v̄, S) = 0. Thus H = H 0 . It follows that ϕ(H) = 0 and ϕ(S \ H) = 0,
hence ϕ(S) = 0, a contradiction. This finishes the proof in this case.
Case 2: F = F2 , the field of two elements.
As H ∪ H 0 6= S, there exists a w ∈ S such that b(w, v) 6= 0 and b(w, v 0 ) 6= 0. As F = F2 ,
this means that b(w, v) = 1 = b(w, v 0 ). Moreover, by our assumptions, ϕ(v̄) = 0 and
ϕ(w) = 0. Consider the linear map
γ:V →V given by γ(x) = x + b(v̄, x)w + b(w, x)v̄.
0 2
Note that
¡ b(w,¢ v̄) = b(w, v)+b(w, v ) = 1+1 = 0. A simple calculation shows that γ = Id
and ϕ γ(x) = ϕ(x) for any x ∈ V , i.e., γ is an isometry. Moreover, γ(v) = v + v̄ = v 0 .
Finally, γ|W0 = Id since w and v̄ are orthogonal to W0 . ¤
Theorem 8.4. (Witt Cancellation Theorem) Let ϕ, ϕ0 be quadratic forms on V and V 0
respectively and ψ, ψ 0 quadratic forms on W and W 0 respectively with rad bψ = 0 = rad bψ0 .
If
ϕ ⊥ ψ ' ϕ0 ⊥ ψ 0 and ψ ' ψ 0
then ϕ ' ϕ0 .
Proof. Let f : ψ → ψ 0 be an isometry. By the Witt Extension Theorem, f extends
to an isometry f˜ : ϕ ⊥ ψ → ϕ0 ⊥ ψ 0 . As fe takes V = W ⊥ to V 0 = (W 0 )⊥ , the result
follows. ¤
Witt Cancellation together with our previous computations allows us to derive the
decomposition that we want.
Theorem 8.5. (Witt Decomposition Theorem) Let ϕ be a quadratic form on V . Then
there exist subspaces V1 and V2 of V such that ϕ = ϕ|rad ϕ ⊥ ϕ|V1 ⊥ ϕ|V2 with ϕ|V1
anisotropic and ϕ|V2 hyperbolic. Moreover, ϕ|V1 and ϕ|V2 are unique up to isometry.
Proof. We know that ϕ = ϕ|rad ϕ ⊥ ϕ|V 0 with ϕV 0 on V 0 unique up to isometry.
Therefore, we can assume that ϕ is regular. Suppose that ϕV 0 is isotropic. By Proposition
7.14, we can split off a subform as an orthogonal summand isometric to the hyperbolic
plane. The desired decomposition follows by induction. As every hyperbolic form is non-
degenerate, the Witt Cancellation Theorem shows the uniqueness of ϕ|V1 up to isometry
hence ϕ|V2 is unique by dimension count. ¤
8.A. Witt equivalence. Using the Witt Decomposition Theorem 8.5, we can define
an equivalence of quadratic forms over a field F .
Definition 8.6. Let ϕ be a quadratic form on V and ϕ = ϕ|rad ϕ ⊥ ϕ|V1 ⊥ ϕ|V2 be
the decomposition in the theorem. The anisotropic form ϕ|V1 , unique up to isometry,
will be denoted ϕan on the space Vϕan and be called the anisotropic part of ϕ. As ϕV2 is
hyperbolic, dim V2 = 2n for some unique non-negative number n. The integer n is called
the Witt index of ϕ and denoted by i0 (ϕ). We say that two quadratic forms ϕ and ψ are
Witt equivalent and write ϕ ∼ ψ if dim rad ϕ = dim rad ψ and ϕan ' ψan . Equivalently,
ϕ ∼ ψ if and only if ϕ ⊥ nH ' ψ ⊥ mH for some n and m.
Note that if ϕ ∼ ψ then ϕK ∼ ψK for any field extension K/F .
Witt cancellation does not hold in general for non-degenerate quadratic forms in char-
acteristic two. We show in the next result, Proposition 8.8, that
(8.7) [a, b] ⊥ hai ' H ⊥ hai
if char F = 2 for all a, b ∈ F with a 6= 0. But [a, b] ' H if and only if [a, b] is isotropic by
Proposition 7.20(iv). Although Witt cancellation does not hold in general in characteristic
two, we can use the following:
Proposition 8.8. Let ρ be a non-degenerate quadratic form of even dimension over

a field F of characteristic 2. Then ρ ⊥ hai ∼ hai for some a ∈ F × if and only if ρ ∼ [a, b]
for some b ∈ F .
Proof. The case char F 6= 2 is easy, so we can assume that char F = 2. Let ϕ =
[a, b] ⊥ hai with a, b ∈ F and a 6= 0. Clearly, ϕ is isotropic and it is non-degenerate as
ϕ|rad bϕ = hai. It follows by Proposition 7.14 that [a, b] ⊥ hai ' H ⊥ hai ∼ hai. Since
ρ ∼ [a, b], we have ρ ⊥ hai ∼ hai.
Conversely, suppose that ρ ⊥ hai ∼ hai for some a ∈ F × . We prove the statement by
induction on n = dim ρ. If n = 0 we can take b = 0. So assume that n > 0. We may also
assume that ρ is anisotropic. By assumption, the form ρ ⊥ hai is isotropic. Therefore,
a = ρ(v) for some v ∈ Vρ , and we can find a decomposition ρ = ρ0 ⊥ [a, d] for some
non-degenerate form ρ0 of dimension n − 2 and b ∈ F . As [a, d] ⊥ hai ' H ⊥ hai by the
first part of the proof, we have
hai ∼ ρ ⊥ hai = ρ0 ⊥ [a, d] ⊥ hai ∼ ρ0 ⊥ hai.
By the induction hypothesis, ρ0 ' [a, c] for some c ∈ F . Therefore by Example 7.24,
ρ = ρ0 ⊥ [a, d] ∼ [a, c] ⊥ [a, d] ' [a, c + d] ⊥ H ∼ [a, c + d]. ¤
Remark 8.9. Let ϕ and ψ be quadratic forms over F .
(1) If ϕ is non-degenerate and anisotropic over F and K/F a purely transcendental exten-
sion then ϕK remains anisotropic by Lemma 7.16. In particular, for any non-degenerate
ϕ, we have i0 (ϕ) = i0 (ϕK ).
(2) Let a ∈ F × . Then ϕ ' aψ if and only if ϕan ' aψan as any form similar to a hyperbolic
form is hyperbolic.
(3) If char F = 2, the quadratic form ϕan may be degenerate. This is not possible if
char F 6= 2.
(4) If char F 6= 2 then every symmetric bilinear form corresponds to a quadratic form,
hence the Witt theorems hold for symmetric bilinear forms in characteristic different from
two.
8.B. Totally isotropic subspaces. Given a regular quadratic form ϕ on V , we

show that every totally isotopic subspace of V lies in a maximal isotropic subspace of V
of dimension equal to the Witt index of ϕ.
Lemma 8.10. Let ϕ be a regular quadratic form on V with W ⊂ V a totally isotropic
subspace of dimension m. If ψ is the quadratic form on W ⊥ /W induced by the restriction
of ϕ on W ⊥ then ϕ ' ψ ⊥ mH.
Proof. As W ∩ rad bϕ ⊂ rad ϕ, the intersection W ∩ rad bϕ is trivial. Thus the map
V → W ∗ by v 7→ lv |W : w 7→ bϕ (v, w) is surjective by Proposition 1.6 and dim W ⊥ =
dim V − dim W . Let W 0 ⊂ V be a subspace mapping isomorphically onto W ∗ . Clearly,
W ∩ W 0 = {0}. Let U = W ⊕ W 0 .
We show the form ϕ|U is hyperbolic. The subspace W ⊕ W 0 is non-degenerate with
respect to bϕ . Indeed let 0 6= v = w + w0 ∈ W ⊕ W 0 . If w0 6= 0 there exists a w0 ∈ W
such that bϕ (w0 , w0 ) 6= 0 hence bϕ (v, w0 ) 6= 0. If w0 = 0, there exists w00 ∈ W 0 such that
bϕ (w, w00 ) 6= 0 hence bϕ (v, w00 ) 6= 0. Thus by Proposition 7.29, the form ϕ|U is isometric
to mH where m = dim W .
By Proposition 7.23, we have ϕ = ϕ|U ⊥ ⊥ ϕ|U ' ϕ|U ⊥ ⊥ mH. As W and U ⊥ are
subspaces of W ⊥ and U ∩ W ⊥ = W , we have W ⊥ = W ⊕ U ⊥ . Thus W ⊥ /W ' U ⊥ and
the result follows. ¤
Proposition 8.11. Let ϕ be a regular quadratic form on V . Then every totally
isotropic subspace of V is contained in a totally isotropic subspace of dimension i0 (ϕ).
Proof. Let W ⊂ V be a totally isotropic subspace of V . We may assume that it
is a maximal totally isotropic subspace. In the notation in the proof of Lemma 8.10, we
have ϕ = ϕ|U ⊥ ⊥ ϕ|U with ϕ|U ' mH where m = dim W . The form ϕ|U ⊥ is anisotropic
by the maximality of W hence must be ϕan by the Witt Decomposition Theorem 8.5. In
particular, dim W = i0 (ϕ). ¤
Corollary 8.12. Let ϕ be a regular quadratic form on V . Then every totally isotropic
subspace W of V has dimension at most i0 (ϕ) with equality if and only if W is a maximal
totally isotropic subspace of V .
Let ρ be a non-degenerate quadratic form and ϕ a subform of ρ. If bϕ is non-degenerate
then ρ = ϕ ⊥ ϕ⊥ hence ρ ⊥ (−ϕ) ∼ ϕ⊥ . However, in general, ρ 6= ϕ ⊥ ϕ⊥ . We do always
have:
Lemma 8.13. Let ρ be a non-degenerate quadratic form of even dimension and ϕ a
regular subform of ρ. Then ρ ⊥ (−ϕ) ∼ ϕ⊥ .
Proof. Let W be the subspace defined by W = {(v, v) | v ∈ Vϕ } of Vρ ⊕ Vϕ . Clearly
W is totally isotropic with respect to the form ρ ⊥ (−ϕ) on Vρ ⊕ Vϕ . By the proof of
Lemma 8.10, we have dim W ⊥ /W = dim Vρ ⊕ Vϕ − 2 dim W = dim Vρ − dim Vϕ . By
Remark 7.11, we also have dim Vϕ⊥ = dim Vρ − dim Vϕ . It follows that the linear map
W ⊥ /W → Vϕ⊥ defined by (v, v 0 ) 7→ v − v 0 is an isometry. On the other hand, by Lemma
8.10, the form on W ⊥ /W is Witt equivalent to ρ ⊥ (−ϕ). ¤
Let V and W be vector spaces over F . Let b be a symmetric bilinear form on W and
ϕ be a quadratic form on V . The tensor product of b and ϕ is the quadratic form b ⊗ ϕ
on W ⊗F V defined by
(8.14) (b ⊗ ϕ)(w ⊗ v) = b(w, w) · ϕ(v)
for all w ∈ W and v ∈ V with the polar form of b ⊗ ϕ equal to b ⊗ bϕ . For example, if
a ∈ F then haib ⊗ ϕ ' aϕ.
Example 8.15. If b is a symmetric bilinear form then ϕb ' b ⊗ h1iq .
Lemma 8.16. Let b be a non-degenerate symmetric bilinear form over F and ϕ a non-
degenerate quadratic form over F . In addition, assume that dim ϕ is even if characteristic
of F is two. Then
(1) The quadratic form b ⊗ ϕ is non-degenerate.
(2) If either ϕ or b is hyperbolic then b ⊗ ϕ is hyperbolic.
Proof. (1): The bilinear form bϕ is non-degenerate by Remark 7.21 and by Remark
7.22 if characteristic of F is not two or two respectively. By Lemma 2.1, the form b ⊗ bϕ
is non-degenerate hence so is b ⊗ ϕ.
(2): Using Proposition 7.29, we see that Vb⊗ϕ contains a totally isotropic space of dimen-
sion 21 dim(b ⊗ ϕ). ¤
As the orthogonal sum of even dimensional non-degenerate quadratic forms over F is

non-degenerate, the isometry classes of even dimensional non-degenerate quadratic forms
over F form a monoid under orthogonal sum. The quotient of the Grothendieck group
of this monoid by the subgroup generated by the image of the hyperbolic plane is called
the quadratic Witt group and will be denoted by Iq (F ). The tensor product of a bilinear
with a quadratic form induces a W (F )-module structure on Iq (F ) by Lemma 8.16.
Remark 8.17. Let ϕ and ψ be two non-degenerate even dimensional quadratic forms
over F . By the Witt Decomposition Theorem 8.5,
ϕ'ψ if and only if ϕ = ψ in Iq (F ) and dim ϕ = dim ψ.
Remark 8.18. Let F → K be a homomorphism of fields. Analogous to Proposition

2.7, this map induces the restriction map
rK/F : Iq (F ) → Iq (K).
It is a group homomorphism. If K/F is purely transcendental, the restriction map is

injective by Lemma 7.16.
Suppose that char F 6= 2. Then we have an isomorphism I(F ) → Iq (F ) given by

b 7→ ϕb . We will use the correspondence b 7→ ϕb to identify bilinear forms in W (F ) with
quadratic forms. In particular, we shall view the class of a quadratic form in the Witt
ring of bilinear forms when char F 6= 2.
9. Quadratic Pfister forms I
As in the bilinear case, there is a special class of forms built from tensor products
of forms. If the characteristic of F is different from two, these forms can be identified
with the bilinear Pfister forms. If the characteristic is two, these forms arise as the tensor
product of a bilinear Pfister form and a binary quadratic form of the type [1, a] and were
first introduced by Baeza (cf. [17]). In general, the quadratic 1-fold Pfister forms are
just the norm forms of a quadratic étale F -algebra and the 2-fold quadratic Pfister forms
are just the reduced norm forms of quaternion algebras. These forms as their bilinear
analogue satisfy the property of being round. In this section, we begin their study.
9. QUADRATIC PFISTER FORMS I 47
9.A. Values and similarities of quadratic forms. Analogously to the symmetric

bilinear case, we study the values that a quadratic form can take as well as the similarity
factors. We begin with some notation.
Definition 9.1. Let ϕ be a quadratic form on V over F . Let
D(ϕ) := {ϕ(v) | v ∈ V, ϕ(v) 6= 0},
the set on nonzero values of ϕ and
G(ϕ) := {a ∈ F × | aϕ ' ϕ},
a group called the group of similarity factors of b. If D(ϕ) = F × , we say that ϕ is
universal. Also set
e
D(ϕ) := D(ϕ) ∪ {0}.
e
We say that elements in D(ϕ) are represented by ϕ.
For example, G(H) = F × (as for bilinear hyperbolic planes) and D(H) = F × . In particu-
lar, if ϕ is an regular isotropic quadratic form over F then ϕ is universal by Proposition
7.14.
The analogous proof of Lemma 1.14 shows:
Lemma 9.2. Let ϕ be a quadratic form. Then
D(ϕ) · G(ϕ) ⊂ D(ϕ).
In particular, if 1 ∈ D(ϕ) then G(ϕ) ⊂ D(ϕ).
The relationship between values and similarities of a symmetric bilinear form and its
associated quadratic form is given by the following:
Lemma 9.3. Let b a symmetric bilinear form on F and ϕ = ϕb . Then
(1) D(ϕ) = D(b).
(2) G(b) ⊂ G(ϕ).
Proof. (1): By definition, ϕ(v) = b(v, v) for all v ∈ V .
¡ ¢ ¡ ¢
(2): Let a ∈ G(b) and λ : b → ab an isometry. Then ϕ λ(v) = b λ(v), λ(v) =
ab(v, v) = aϕ(v) for all v ∈ V . ¤
A quadratic form is called round if G(ϕ) = D(ϕ). In particular, if ϕ is round then

D(ϕ) is a group. For example, any hyperbolic form is round.
A basic example of round forms arises from quadratic F -algebras (Cf. §98.B):
Example 9.4. Let K be a quadratic F -algebra. Then there exists an involution on
K denoted by x 7→ x̄ and a quadratic norm form ϕ = N given by x 7→ xx̄ (cf. §98.B). We
have ϕ(xy) = ϕ(x)ϕ(y) for all x, y ∈ K. If x ∈ K with ϕ(x) 6= 0 then x ∈ K × . Hence the
map K → K given by multiplication by x is an F -isomorphism and ϕ(x) ∈ G(ϕ). Thus
D(ϕ) ⊂ G(ϕ). As 1 ∈ D(ϕ), we have G(ϕ) ⊂ D(ϕ). In particular, ϕ is round.
Let K be a quadratic étale F -algebra. So K = Fa for some a ∈ F . Denote the norm

form N of Fa in Example 9.4 by hha]] and called it a quadratic 1-fold Pfister form. In
particular, it is round. Explicitly, we have:
Example 9.5. For Fa a quadratic étale F algebra, we have
(1) (Cf. Example 98.3.) If char F 6= 2 then Fa = F [t]/(t2 − a) with a ∈ F × and the
quadratic form hha]] = h1, −aiq ' hhaiib ⊗ h1iq is the norm form of Fa .
(2) (Cf. Example 98.4.) If char F = 2 then Fa = F [t]/(t2 + t + a) with a ∈ F and the
quadratic form hha]] = [1, a] is the norm form of Fa . In particular, hha]] ' hhx2 + x + a]]
for any x ∈ F .
9.B. Quadratic Pfister forms and round forms. Let n ≥ 1. A quadratic form
isometric to a quadratic form of the type
hha1 , . . . , an ]] := hha1 , . . . , an−1 iib ⊗ hhan ]]
for some a1 , . . . , an−1 ∈ F × and an ∈ F (with an 6= 0 if char F 6= 2) is called a quadratic
n-fold Pfister form. It is convenient to call the form isometric to h1iq a 0-fold Pfister form.
Every quadratic n-fold Pfister form is non-degenerate by Lemma 8.16. We let
Pn (F ) := {ϕ | ϕ a quadratic n-fold Pfister form}
[
P (F ) := Pn (F )
GPn (F ) := {aϕ | a ∈ F × , ϕ a quadratic n-fold Pfister form}
[
GP (F ) := GPn (F ).
Forms in GPn (F ) are called general quadratic n-fold Pfister forms.
If char F = 6 2, the form hha1 , . . . , an ]] is the associated quadratic form of the bi-
linear Pfister form hha1 , . . . , an iib by Example 9.5 (1). We shall also use the notation
hha1 , . . . , an ii for the quadratic Pfister form hha1 , . . . , an ]] in this case.
The class of an n-fold Pfister form (n > 0) belongs to
Iqn (F ) := I n−1 (F ) · Iq (F ).
As [a, b] = a[1, ab] for all a, b ∈ F with a 6= 0, every non-degenerate binary quadratic
form is a general 1-fold Pfister form. In particular, GP1 (F ) generates Iq (F ). It follows
that GPn (F ) generates Iqn (F ) as an abelian group. In fact, as
(9.6) ahhb, c]] = hhab, c]] − hha, c]]
for all a, b ∈ F × and c ∈ F (with c 6= 0 if char F 6= 2), Pn (F ) generates Iqn (F ) as an
abelian group for n > 1.
Note that in the case that char F 6= 2, under the identification of I(F ) with Iq (F ), the
group I n (F ) corresponds to Iqn (F ) and a bilinear Pfister form hha1 , . . . , an iib corresponds
to the quadratic Pfister form hha1 , . . . , an ii.
Using the material in §98.E, we have the following example.
9. QUADRATIC PFISTER FORMS I 49
Example 9.7. Let A be a quaternion F -algebra.

µ ¶
a, b
(1) (Cf. Example 98.11.) Suppose that char F 6= 2. If A = then the reduced
F
quadratic norm form is equal to the quadratic form h1, −a, −b, abi = hha, bii.
· ¸
a, b
(2) (Cf. Example 98.12.) Suppose that char F = 2. If A = then the reduced
F
quadratic norm form is equal to the quadratic form [1, ab] ⊥ [a, b]. This form is hyperbolic
if a = 0 and is isomorphic to h1, aib ⊗ [1, ab] = hha, ab]] otherwise.
Example 9.8. Let L/F be a separable quadratic field extension and Q = (L/F, b), i.e.,
Q = L ⊕ Lj a quaternion F -algebra with j 2 = b ∈ F × (cf. §98.E). For any q = l + l0 j ∈ Q,
we have NrdQ (q) = NL (l) − b NL (l0 ). Therefore, NrdQ ' hhbii ⊗ NL .
Proposition 9.9. Let ϕ be a round quadratic form and a ∈ F × . Then

(1) The form hhaii ⊗ ϕ is also round.
(2) If ϕ is regular then the following are equivalent:
(i) hhaii ⊗ ϕ is isotropic.
(ii) hhaii ⊗ ϕ is hyperbolic.
(iii) a ∈ D(ϕ).
Proof. Set ψ = hhaii ⊗ ϕ.
(1): Since 1 ∈ D(ϕ), it suffices to prove that D(ψ) ⊂ G(ψ). Let c be a nonzero value of ψ.
e
Write c = x − ay for some x, y ∈ D(ϕ). If y = 0, we have c = x¡ ∈ D(ϕ)
¢ = G(ϕ) ⊂ G(ψ).
Similarly, y ∈ G(ψ) if x = 0 hence c = −ay ∈ G(ψ) as −a ∈ G hhaii ⊂ G(ψ).
Now suppose that x and y are nonzero. Since ϕ is round, x, y ∈ G(ϕ) and therefore
ψ = ϕ ⊥ (−aϕ) ' ϕ ⊥ (−ayx−1 )ϕ = hhayx−1 ii ⊗ ϕ.
¡ ¢
By Example 1.15, we know that 1 − ayx−1 ∈ G hhayx−1 ii ⊂ G(ψ). Since x ∈ G(ϕ) ⊂
G(ψ), we have c = (1 − ayx−1 )x ∈ G(ψ).
(2): (i) ⇒ (iii): If ϕ is isotropic then ϕ is universal by Proposition 7.14. So suppose that
ϕ is anisotropic. Since ψ = ϕ ⊥ (−aϕ) is isotropic, there exist x, y ∈ D(ϕ) such that
x − ay = 0. Therefore a = xy −1 ∈ D(ϕ) as D(ϕ) is closed under multiplication.
(iii) ⇒ (ii): As ϕ is round, a ∈ D(ϕ) = G(ϕ) and hhaii ⊗ ϕ is hyperbolic.
(ii) ⇒ (i) is trivial. ¤
Corollary 9.10. Quadratic Pfister forms are round.
Corollary 9.11. A quadratic Pfister form is either anisotropic or hyperbolic.
Proof. Suppose that ψ is an isotropic quadratic n-fold Pfister form. If n = 1 the
result follows by Proposition 7.20(iv). So assume that n > 1. Then ψ ' hhaii ⊗ ϕ for a
quadratic Pfister form ϕ and the result follows by Proposition 9.9. ¤
Let char F = 2. We need another characterization of hyperbolic Pfister forms in this
case. Let ℘ : F → F be the Artin-Schreier map defined by ℘(x) = x2 + x. (Cf. §98.B.)
For a quadratic 1-fold Pfister form we have hhd]] is hyperbolic if and only if d ∈ im(℘) by
Example 98.4. More generally, we have:
Lemma 9.12. Let b be an anisotropic bilinear Pfister form and d ∈ F . Then b ⊗ hhd]]
e 0 ).
is hyperbolic if and only if d ∈ im(℘) + D(b
Proof. Suppose that b ⊗ hhd]] is hyperbolic and therefore isotropic. Let {e, f } be
the standard basis of hhd]]. Let v ⊗ e + w ⊗ f be an isotropic vector of b ⊗ hhd]] where
v, w ∈ Vb . We have a + b + cd = 0 where a = b(v, v), b = b(v, w), and c = b(w, w).
As b is anisotropic, we have w 6= 0, i.e., c 6= 0. Suppose first that v = sw for some
s ∈ F . Then 0 = a + b + cd = c(s2 + s + d), hence d = s2 + s ∈ im(℘).
Now suppose that v and w generate a 2-dimensional subspace W of Vb . The determi-
nant of b|W is equal to xF ×2 where x = b2 + bc + c2 d. Hence b|W ' chhxii by Example
1.11. As c ∈ D(b) = G(b) by Corollary 6.2, the form hhxii is isometric to a subform of
b. By the Bilinear Witt Cancellation Theorem 1.29, we have hxi is a subform of b0 , i.e.,
x ∈ D(b0 ). Hence (b/c)2 + (b/c) + d = x/c2 ∈ D(b0 ) and therefore d ∈ im(℘) + D(b0 ).
Conversely, let d = x + y where x ∈ im(℘) and y ∈ D(b e 0 ). If y = 0 then hhd]] is
hyperbolic hence so is b ⊗ hhd]]. So suppose that y 6= 0. By Lemma 6.11 there is a bilinear
Pfister form c such that b ' c ⊗ hhyii. Therefore, b ⊗ hhd]] ' c ⊗ hhy, dii is hyperbolic as
hhy, d]] ' hhy, y]] by Example 98.4 which is hyperbolic. ¤
9.C. Annihilators. If ϕ is a non-degenerate quadratic form over F then the anni-

hilator of ϕ in W (F )
annW (F ) (ϕ) := {c ∈ W (F ) | c · ϕ = 0}
is an ideal. When ϕ is a Pfister form this ideal has the structure that we had when ϕ was
a bilinear anisotropic Pfister form. Indeed the same proof yielding Proposition 6.22 and
Corollary 6.23 shows:
Theorem 9.13. Let ϕ be anisotropic quadratic Pfister form. Then annW (F ) (ϕ) is
generated by binary symmetric bilinear forms hhxiib with x ∈ D(ϕ).
As in the bilinear case, if ϕ is 2-dimensional, we obtain stronger results. Indeed the
same proofs for the corresponding results show
Lemma 9.14. (Cf. Lemma 6.24.) Let ϕ be a binary anisotropic quadratic form over
F and c an anisotropic bilinear form over F such that c ⊗ ϕ is isotropic. Then c ' d ⊥ e
for some binary bilinear form d annihilated by ϕ and bilinear form e over F .
Proposition 9.15. (Cf. Proposition 6.25.) Let ϕ be a binary anisotropic quadratic
form over F and c an anisotropic bilinear form over F . Then there exist bilinear forms
c1 and c2 over F such that c ' c1 ⊥ c2 with c2 ⊗ ϕ anisotropic and c1 ' d1 ⊥ · · · ⊥ dn
where each di is a binary bilinear form annihilated by ϕ. In particular, − det di ∈ D(ϕ)
for each i.
Corollary 9.16. (Cf. Corollary 6.26.) Let ϕ be a binary anisotropic quadratic form
over F and c an anisotropic bilinear form over F annihilated by b. Then c ' d1 ⊥ · · · ⊥ dn
for some binary bilinear forms di annihilated by b for i = 1, . . . , n.
10. TOTALLY SINGULAR FORMS 51
10. Totally singular forms
Totally singular forms in characteristic different from two are zero forms but in char-
acteristic two they become interesting. In this section, we look at totally singular forms
in characteristic two. In particular, throughout most of this section, char F = 2.
Let char F = 2. Let ϕ be a quadratic form over F . Then ϕ is totally singular form
if and only if it is diagonalizable. Moreover, if this is the case, then every basis of Vϕ is
orthogonal by Remark 7.25 and D(ϕ) e is a vector space over the field F 2 .
¡ ¢
We investigate the F -subspace (D e ϕ) 1/2 of F 1/2 . Define an F -linear map
¡ ¢1/2 p
e
f : Vϕ → D(ϕ) given by f (v) = ϕ(v).
¡ ¢1/2
Then f is surjective and ker(f ) = rad ϕ. Let ϕ e be the quadratic form on D(ϕ) e over
√
F defined by ϕ( e a) = a. Clearly ϕ e is anisotropic. Consequently, if ϕ̄ is the quadratic
form induced on Vϕ / rad ϕ by ϕ then f induces an isometry between ϕ̄ and ϕ. e Moreover
ϕ
e ' ϕan . Therefore, if char F = 2, the correspondence ϕ 7→ D(ϕ)e gives rise to a bijection
Isometry classes of totally singular ∼ Finite dimensional

−→
anisotropic quadratic forms F 2 -subspaces of F
Moreover, for any totally singular quadratic form ϕ, we have

e
dim ϕan = dim D(ϕ)
and if ϕ and ψ are two totally singular quadratic forms then
ϕ'ψ if and only if D(ϕ) = D(ψ) and dim ϕ = dim ψ.
e ⊥ ψ) = D(ϕ)
We also have D(ϕ e e
+ D(ψ).
Example 10.1. If F is a separably closed field of characteristic two, the anisotropic
quadratic forms are diagonalizable hence totally singular.
Note that if b is an alternating bilinear form and ψ is a totally singular quadratic form
then b ⊗ ψ = 0. It follows that the tensor product of totally singular quadratic forms
e ⊗ ψ)
ϕ ⊗ ψ := c ⊗ ψ is well-defined where c is a bilinear form with ϕ = ϕc . The space D(ϕ
2
is spanned by D(ϕ) · D(ψ) over F .
Proposition 10.2. Let char F = 2. If ϕ is a totally singular quadratic form then
G(ϕ) = {a ∈ F × | aD(ϕ) ⊂ D(ϕ)}.
Proof. The inclusion “⊂” follows from Lemma 9.2. Conversely, let a ∈ F × satisfy
¡ ¢1/2 ¡ ¢1/2 √
e
aD(ϕ) ⊂ D(ϕ). Then the F -linear map g : D(ϕ) e
→ D(ϕ) defined by g(b) = a b
is an isometry between ϕ
e and aϕ.
e Therefore a ∈ G(ϕ)
e = G(ϕ). ¤
e
It follows from Proposition 10.2 that G(ϕ) := G(ϕ) ∪ {0} is a subfield of F containing
2 e e
F and D(ϕ) is a vector space over G(ϕ).
It is also convenient to introduce a variant of the notion of Pfister forms in all charac-
teristics. A quadratic form ϕ is called a quasi-Pfister form if there exists a bilinear Pfister
form b with ϕ ' ϕb , i.e., ϕ ' hha1 , . . . , an iib ⊗ h1iq . for some a1 , . . . , an ∈ F × . Denote
hha1 , . . . , an iib ⊗ h1iq by hha1 , . . . , an iiq .
If char F 6= 2 then the classes of quadratic Pfister and quasi-Pfister forms coincide. If
char F = 2 every quasi-Pfister form is totally singular. Quasi-Pfister forms have some
properties similar to those for quadratic Pfister forms.
Corollary 10.3. Quasi-Pfister forms are round.
Proof. Let b be a bilinear Pfister form. As h1iq is a round quadratic form, the
quadratic form b ⊗ h1iq is round by Proposition 9.9. ¤
Remark 10.4. Let char F = 2. Let ρ = hha1 , . . . , an iiq be an anisotropic quasi-Pfister
e
form. Then D(ρ) is equal to the field F 2 (a1 , . . . , an ) of degree 2n over F 2 . Conversely
every field K such that F 2 ⊂ K ⊂ F with [K : F 2 ] = 2n is generated by n elements and
e
therefore K = D(ρ) for an anisotropic n-fold quasi-Pfister form ρ. Thus we get a bijection
Isometry classes of anisotropic Fields K with F 2 ⊂ K ⊂ F

'
n-fold quasi-Pfister forms and [K : F 2 ] = 2n
Let ϕ be an anisotropic totally singular quadratic form. Then K = G(ϕ) e is a field with
e e 2 e
K · D(ϕ) ⊂ D(ϕ). We have [K : F ] < ∞ and D(ϕ) is a vector space over K. Let
e
{b1 , . . . , bm } be a basis of D(ϕ) over K and set ψ = hb1 , . . . , bm iq . Choose an anisotropic
n-fold quasi-Pfister form ρ such that D(ρ) e e
= G(ϕ). e
As D(ϕ) is the vector space spanned
2
by K · D(ψ) over F , we have ϕ ' ρ ⊗ ψ. In fact, ρ is the largest quasi-Pfister divisor of
ϕ, i.e., quasi-Pfister form of maximal dimension such that ϕ ' ρ ⊗ ψ.
11. The Clifford algebra
To each quadratic form ϕ, one associates a Z/2Z-graded algebra by factoring the

tensor algebra on Vϕ by the relation ϕ(v) = v 2 . This algebra, called the Clifford algebra
generalizes the exterior algebra. In this section, we study the basic properties of Clifford
algebras.
Let ϕ be a quadratic form on V over F . Define` the⊗n Clifford algebra of ϕ to be the
factor algebra C(ϕ) of the tensor algebra T (V ) = n≥0 V modulo the ideal I generated
by (v ⊗ v) − ϕ(v) for all v ∈ V . We shall view vectors in V as elements of C(ϕ) via the
natural F -linear map V → C(ϕ). Note that v 2 = ϕ(v) in C(ϕ) for every v ∈ V . The
Clifford algebra of ϕ has a natural Z/2Z-grading
C(ϕ) = C0 (ϕ) ⊕ C1 (ϕ)
11. THE CLIFFORD ALGEBRA 53
as I is homogeneous if degree is viewed modulo two. The subalgebra C0 (ϕ) is called the
even Clifford algebra of ϕ. We have dim C(ϕ) = 2dim ϕ and dim C0 (ϕ) = 2dim ϕ−1 . If K/F
is a field extension, C(ϕK ) = C(ϕ)K and C0 (ϕK ) = C0 (ϕ)K .
Lemma 11.1. Let ϕ be a quadratic form on V over F with polar form b. Let v, w ∈ V .
Then b(v, w) = vw + wv in C(ϕ). In particular, v and w are orthogonal if and only if
vw = −wv in C(ϕ).
Proof. This follows from the polar identity. ¤
Example 11.2. (1) The V Clifford algebra of the zero quadratic form on V coincides
with the exterior algebra V .
¡ ¢
(2) C0 hai = F .
¡ ¢
(3)
µ If¶char F =
6 2 then the Clifford algebra of the quadratic form ha, bi is C ha, bi =
a, b ¡ ¢ ¡ ¢
and C0 ha, bi = F−ab . In particular, C0 hhbii = Fb .
F
· ¸
¡ ¢ a, b ¡ ¢ ¡ ¢
(4) If char F = 2 then C [a, b] = and C0 [a, b] = Fab . In particular, C0 hhb]] =
F
Fb .
By construction, the Clifford algebra satisfies the following universal property:
For any F -algebra A and any F -linear map f : V → A satisfying f (v)2 = ϕ(v) for
all v ∈ V , there exists a unique F -algebra homomorphism f˜ : C(ϕ) → A satisfying
f˜(v) = f (v) for all v ∈ V .
Example 11.3. Let C(ϕ)op denote the Clifford algebra of ϕ with the opposite multipli-
cation. The canonical linear map V → C(ϕ)op extends to an involution : C(ϕ) → C(ϕ)
given by the algebra isomorphism C(ϕ) → C(ϕ)op . Note that if x = v1 v2 · · · vn then
x̄ = vn · · · v2 v1 .
Proposition 11.4. Let ϕ be a quadratic form on V over F and let a ∈ F × . Then
(1) C0 (aϕ) ' C0 (ϕ), i.e., the even Clifford algebras of similar quadratic forms are
isomorphic.
(2) Let ϕ = hai ⊥ ψ. Then C0 (ϕ) ' C(−aψ).
Proof. (1): Set K = F [t]/(t2 − a) = F ⊕ F t̄ with t̄ the image of t in K. Since (v ⊗
t̄)2 = ϕ(v) ⊗ t̄2 = aϕ(v) ⊗ 1 in C(ϕ)K = C(ϕ) ⊗F K, there is an F -algebra homomorphism
α : C(aϕ) → C(ϕ)K taking v ∈ V to v ⊗ t̄ by the universal property of the Clifford algebra
aϕ. Since
(v ⊗ t̄)(v 0 ⊗ t̄) = vv 0 ⊗ t̄2 = avv 0 ⊗ 1 ∈ C(ϕ) ⊂ C(ϕ)K ,
the map α restricts to an F -algebra homomorphism C0 (aϕ) → C0 (ϕ). As this map is
clearly a surjective map of algebras of the same dimension, it is an isomorphism.
(2): Let V = F v ⊕ W with ϕ(v) = a and W ⊂ (F v)⊥ . Since
(vw)2 = −v 2 w2 = −ϕ(v)ψ(w) = −aψ(w)
for every w ∈ W , the map W → C0 (ϕ) defined by w 7→ vw extends to an F -algebra
∼
isomorphism C(−aψ) → C0 (ϕ) by the universal property of Clifford algebras. ¤
Let ϕ be a quadratic form on V over F . Applying the universal property of Clifford

algebras to the natural linear map V → V / rad ϕ → C(ϕ̄), where ϕ̄ is the induced
quadratic form on V / rad ϕ, we get a surjective F -algebra homomorphism C(ϕ) → C(ϕ̄)
with kernel (rad ϕ)C(ϕ). Consequently, we get canonical isomorphisms
C(ϕ̄) ' C(ϕ)/(rad ϕ)C(ϕ),
C0 (ϕ̄) ' C0 (ϕ)/(rad ϕ)C1 (ϕ).
Example 11.5. Let ϕ = H(W ) be the hyperbolic form on the vector space V =
W ⊕ W ∗ with W a nonzero vector space. Then
^
C(ϕ) ' EndF ( W ),
V
where the exterior algebra W of V is considered as a vector space (cf. [87, Prop. 8.3]).
Moreover, ^ ^
C0 (ϕ) = EndF ( W ) × EndF ( W ),
0 1
where
^ 2i
M^ ^ M 2i+1
^
W := W and W := W.
0 1
i≥0 i≥0
In particular, C(ϕ) is a split central simple F -algebra and the center of C0 (ϕ) is the split
quadratic étale F -algebra F × F . Note also that the natural F -linear map V → C(ϕ) is
injective.
Proposition 11.6. Let ϕ be a quadratic form over F .
(1) If dim ϕ ≥ 2 is even then the following conditions are equivalent:
(a) ϕ is non-degenerate.
(b) C(ϕ) is central simple.
(c) C0 (ϕ) is separable with center Z(ϕ), an étale quadratic algebra.
(2) If dim ϕ ≥ 3 is odd then the following conditions are equivalent:
(a) ϕ is non-degenerate.
(b) C0 (ϕ) is central simple.
Proof. We may assume that F is algebraically closed. Suppose first that ϕ is non-
degenerate and even dimensional. Then ϕ is hyperbolic; and, by Example 11.5, the algebra
C(ϕ) is a central simple F -algebra and C0 (ϕ) is a separable F -algebra whose center is the
split quadratic étale F -algebra F × F .
Conversely, suppose that the even Clifford algebra C0 (ϕ) is separable or C(ϕ) is central
simple. Let v ∈ rad ϕ. The ideals I = vC1 (ϕ) in C0 (ϕ) and J = vC(ϕ) in C(ϕ) satisfy
I 2 = 0 = J 2 . Consequently, I = 0 or J = 0 as C0 (ϕ) is semi-simple or C(ϕ) is central
simple and therefore v = 0 thus rad ϕ = 0. Thus ϕ is non-degenerate.
Now suppose that dim ϕ is odd. Write ϕ = hai ⊥ ψ for some a ∈ F and an even
dimensional form ψ. Let v ∈ Vϕ be a nonzero vector satisfying ϕ(v) = a with v orthogonal
to Vψ . If ϕ is non-degenerate then a 6= 0 and ψ is non-degenerate. It follows from
Proposition 11.4(2) and the first part of the proof that the algebra C0 (ϕ) ' C(−aψ) is
central simple.
12. BINARY QUADRATIC FORMS AND QUADRATIC ALGEBRAS 55
Conversely, suppose that the algebra C0 (ϕ) is central simple. As dim ϕ ≥ 3, the
subspace I := vC1 (ϕ) of C0 (ϕ) is nonzero. If a = 0 then I is a nontrivial ideal of C0 (ϕ),
a contradiction to the simplicity of C0 (ϕ). Thus a 6= 0 and by Proposition 11.4(2),
C0 (ϕ) ' C(−aψ). Hence by the first part of the proof, the form ψ is non-degenerate.
Therefore, ϕ is also non-degenerate. ¤
Lemma 11.7. Let ϕ be a non-degenerate quadratic form of positive even dimension.
Then yx = x̄y for every x ∈ Z(ϕ) and y ∈ C1 (ϕ).
Proof. Let v ∈ Vϕ be an anisotropic vector hence a unit in C(ϕ). Since conjugation
by v on C(ϕ) stabilizes C0 (ϕ), it stabilizes the center of C0 (ϕ), i.e., vZ(ϕ)v −1 = Z(ϕ). As
C(ϕ) is a central algebra, conjugation by v induces a nontrivial automorphism on Z(ϕ)
given by x 7→ x̄ otherwise C1 (ϕ) = C0 (ϕ)v and therefore the full algebra C(ϕ) would
commute with Z(ϕ). Thus vx = x̄v for all x ∈ Z(ϕ). Let y ∈ C1 (ϕ). Writing y in
the form y = zv for some z ∈ C0 (ϕ), we have yx = zvx = z x̄v = x̄zv = x̄y for every
x ∈ Z(ϕ). ¤
Corollary 11.8. Let ϕ be a non-degenerate quadratic form of positive even dimen-
sion. If a is a norm for the quadratic étale algebra Z(ϕ) then C(aϕ) ' C(ϕ).
Proof. Let x ∈ Z(ϕ) satisfy N(x) = a. By Lemma 11.7, we have (vx)2 = N(x)v 2 =
aϕ(v) in C(ϕ) for every v ∈ V . By the universal property of the Clifford algebra aϕ, there
is an algebra homomorphism α : C(aϕ) → C(ϕ) mapping v to vx. Since both algebras
are simple of the same dimension, α is an isomorphism. ¤
12. Binary quadratic forms and quadratic algebras
In appendices §98.E and §98.B, we review the theory of quadratic and quaternion
algebras. In this section, we study the relationship between these algebras and quadratic
forms.
If A is a quadratic F -algebra, we let T rA and NA denote the trace form and the
quadratic norm form of A respectively. (cf. §98.B). Note that NA is a binary form
representing 1.
Conversely, if ϕ is a binary quadratic form over F then the even Clifford algebra C0 (ϕ)
is a quadratic F -algebra. We have defined two maps
Quadratic −→ Binary quadratic

F -algebras ←− forms representing 1
Proposition 12.1. The above two maps induce a bijection on the set of isomorphism
classes of quadratic F -algebras and the set of isometry classes of binary quadratic forms
representing one. Under this bijection, we have:
(1) Quadratic étale algebras correspond to non-degenerate binary forms.
(2) Quadratic fields correspond to anisotropic binary forms.
(3) Semisimple algebras correspond to regular binary quadratic forms.
Proof. Let A be a quadratic F -algebra. We need to show that A ' C0 (NA ). We have
C1 (NA ) = A. Therefore, the map α : A → C0 (NA ) defined by x 7→ 1·x (where dot denotes
the product in the Clifford algebra) is an F -linear isomorphism. We shall show that α is
an algebra isomorphism, i.e., (1 · x) · (1 · y) = 1 · xy for all x, y ∈ A. The equality holds if
x ∈ F or y ∈ F . Since A is 2-dimensional over F , it is suffices to check the equality when
x = y and it does not lie in F . We have 1 · x + x · 1 = NA (x + 1) − NA (x) − NA (1) = TrA (x),
so
(1 · x) · (1 · x) = (1 · x) · (TrA (x) − x · 1) = 1 · TrA (x)x − 1 · NA (x) = 1 · x2
as needed.
Conversely, let ϕ be a binary quadratic form on V representing 1. We shall show that
the norm form for the quadratic F -algebra C0 (ϕ) is isometric to ϕ. Let v0 ∈ V be a
vector satisfying ϕ(v0 ) = 1. Let f : V → C0 (ϕ) be the F -linear isomorphism defined by
f (v) = v · v0 . The quadratic equation (98.2) for v · v0 ∈ C0 (ϕ) in §98.B becomes
(v · v0 )2 = v · (b(v, v0 ) − v · v0 ) · v0 = b(v, v0 )(v · v0 ) − ϕ(v)
so NC0 (ϕ) (v · v0 ) = ϕ(v) hence
¡ ¢
NC0 (ϕ) f (v) = NC0 (ϕ) (v · v0 ) = ϕ(v),
i.e., f is an isometry of ϕ with the norm form of C0 (ϕ) as needed.
In order to prove that quadratic étale algebras correspond to non-degenerate binary
forms it is sufficient to assume that F is algebraically closed. Then a quadratic étale
algebra A is isomorphic to F × F and therefore NA ' H. Conversely, by Example 11.5,
C0 (H) ' F × F .
If a quadratic F -algebra A is a field, then obviously the norm form NA is anisotropic.
Conversely, if NA is anisotropic, then for every nonzero a ∈ A we have aā = NA (a) 6= 0,
therefore a is invertible, i.e., A is a field.
Statement (3) follows from Statements (1) and (2), since a quadratic F -algebra is
semisimple if and only if it is either a field or F × F ; and a binary quadratic form is
regular if and only if it is anisotropic or hyperbolic. ¤
Corollary 12.2. (1) Let A and B be quadratic F -algebras. Then A and B are
isomorphic if and only if the norm forms NA and NB are isometric.
(2) Let ϕ and ψ be nonzero binary quadratic forms. Then ϕ and ψ are similar if and
only if the even Clifford algebras C0 (ϕ) and C0 (ψ) are isomorphic.
Corollary 12.3. Let ϕ be an anisotropic binary quadratic form and let K/F be a
quadratic field extension. Then ϕK is isotropic if and only if K ' C0 (ϕ).
Proof. By Proposition 12.1, the form ϕK is isotropic if and only if the 2-dimensional
even Clifford K-algebra C0 (ϕK ) = C0 (ϕ) ⊗ K is not a field. The later is equivalent to
K ' C0 (ϕ). ¤
We now consider the relationship between quaternion and Clifford algebras.

Proposition 12.4. Let Q be a quaternion F -algebra and ϕ the reduced norm quadratic
form of Q. Then C(ϕ) ' M2 (Q).
13. THE DISCRIMINANT 57
µ ¶
0 x
Proof. For every x ∈ Q, let mx be the matrix in M2 (Q). Since m2x =
x̄ 0
xx̄ = Nrd(x) = ϕ(x), the F -linear map Q → M2 (Q) defined by x 7→ mx extends to
an F -algebra homomorphism α : C(ϕ) → M2 (Q) by the universal property of Clifford
algebras. As C(ϕ) is a central simple algebra of dimension 16 = dim M2 (Q), the map α
is an isomorphism. ¤
Corollary 12.5. Two quaternion algebras are isomorphic if and only if their reduced
norm quadratic forms are isometric. In particular, a quaternion algebra is split if and
only if its reduced norm quadratic form is hyperbolic.
Exercise 12.6. Let Q be a quaternion F -algebra and let ϕ0 be the restriction of
the reduced norm quadratic form to the space Q0 of pure quaternions. Prove that ϕ0 is
non-degenerate and C0 (ϕ0 ) is isomorphic to Q. Conversely, prove that any 3-dimensional
non-degenerate quadratic form ψ is similar to the restriction of the reduced norm quadratic
form of C0 (ψ) to the space of pure quaternions. Therefore, there is a natural bijection
between the set of isomorphism classes of quaternion algebras over F and the set of
similarity classes of 3-dimensional non-degenerate quadratic forms over F .
13. The discriminant
A major objective is to define sufficiently many invariants of quadratic forms. The

first and simplest such invariant is the dimension. In this section, using quadratic étale
algebras, we introduce a second invariant, the discriminant, of a non-degenerate quadratic
form.
Let ϕ be a non-degenerate quadratic form over F of positive even dimension. The
center Z(ϕ) of C0 (ϕ) is a quadratic étale F -algebra. The class of Z(ϕ) in Ét2 (F ), the
group of isomorphisms classes of quadratic étale F -algebras under the operation · induced
by ? (cf. §98.B), is called the discriminant of ϕ and will be denoted by disc(ϕ). Define
the discriminant of the zero form to be trivial.
¡ ¢
Example
¡ ¢ 13.1. By Example 11.2, we have disc ha, bi = F−ab if char F 6= 2 and
disc [a, b] = Fab if char F = 2. It follows from Example 11.5 that the discriminant of a
hyperbolic form is trivial.
The discriminant is a complete invariant for the similarity class of a non-degenerate
binary quadratic form, i.e.,
Proposition 13.2. Two non-degenerate binary quadratic forms are similar if and
only if their discriminants are equal.
Proof. Let disc(ϕ) = disc(ψ), i.e., C0 (ϕ) ' C0 (ψ). Write ϕ = aϕ0 and ψ = bψ 0 ,
where ϕ0 and ψ 0 both represent 1. By Proposition 12.1, the forms ϕ0 and ψ 0 are the
norm forms for C0 (ϕ0 ) = C0 (ϕ) and C0 (ψ 0 ) = C0 (ψ) respectively. Since these algebras are
isomorphic, we have ϕ0 ' ψ 0 . ¤
Corollary 13.3. A non-degenerate binary quadratic form ϕ is hyperbolic if and only
if disc(ϕ) is trivial.
Lemma 13.4. Let ϕ and ψ be non-degenerate quadratic forms of even dimension over
F . Then disc(ϕ ⊥ ψ) = disc(ϕ) · disc(ψ).
¡ ¢
¡ Proof. The even
¢ Clifford algebra C 0 (ϕ ⊥ ψ) coincides with C0 (ϕ) ⊗F C 0 (ψ) ⊕
C1 (ϕ) ⊗F C1 (ψ) and contains Z(ϕ) ⊗F Z(ψ). By Lemma 11.7, we have yx = x̄y for
every x ∈ Z(ϕ) and y ∈ C1 (ϕ). Similarly, wz = z̄t for every z ∈ Z(ψ) and w ∈ C1 (ψ).
Therefore, the center of C0 (ϕ ⊥ ψ) coincides with the subalgebra Z(ϕ) ? Z(ψ) of all stable
elements of Z(ϕ) ⊗F Z(ψ) under the automorphism x ⊗ y 7→ x̄ ⊗ ȳ. ¤
Example 13.5. (1) Let char F =
6 2. Then
¡ ¢
disc ha1 , a2 , . . . , a2n i = Fc
where c = (−1)n a1 a2 . . . a2n . For this reason, the discriminant is often called the signed
determinant when the characteristic of F is different from two.
(2) Let char F = 2. Then
¡ ¢
disc [a1 , b1 ] ⊥ · · · ⊥ [an , bn ] = Fc
where c = a1 b1 + · · · + an bn . The discriminant in the characteristic two case is often called
the Arf invariant.
Proposition 13.6. Let ρ be a non-degenerate quadratic form over F . If disc(ρ) = 1
and ρ ⊥ hai ∼ hai for some a ∈ F × then ρ ∼ 0.
Proof. We may assume that the characteristic¡ ¢ of F is two. By Proposition 8.8, we
have ρ ∼ [a, b] for some b ∈ F . Therefore disc [a, b] is trivial and [a, b] ∼ 0. ¤
It follows from Lemma 13.4 and Example 11.5 that the map
e1 : Iq (F ) → Ét2 (F )
taking a form ϕ to disc(ϕ) is a well-defined group homomorphism.
The analogue of Proposition 4.13 is true, viz.,
Theorem 13.7. The homomorphism e1 is surjective with kernel Iq2 (F ).
Proof. The surjectivity follows from Example 13.1. Since similar forms ¡ have isomor-
¢
×
phic even Clifford algebras, for any¡ ϕ ∈ ¢Iq (F ) and a ∈ F , we have e1 hh−aii · ϕ =
e1 (ϕ) + e1 (−aϕ) = 0. Therefore, e1 Iq2 (F ) = 0.
Let ϕ ∈ Iq (F ) be a form with trivial discriminant. We show by induction on dim ϕ
that ϕ ∈ Iq2 (F ). The case dim ϕ = 2 follows from Corollary 13.3. Suppose that dim ϕ ≥ 4.
Write ϕ = ρ ⊥ ψ with ρ a binary form. Let a ∈ F × be chosen so that the form ϕ0 = aρ ⊥ ψ
is isotropic. Then the class of ϕ0 in Iq (F ) is represented by a form of dimension less than
dim ϕ. As disc(ϕ0 ) = disc(ϕ) is trivial, ϕ0 ∈ Iq2 (F ) by induction. Since ρ ≡ aρ mod Iq2 (F ),
ϕ also lies in Iq2 (F ). ¤
Remark 13.8. One can also define a discriminant like invariant for all non-degenerate
quadratic forms. Let ϕ be a non-degenerate quadratic form. Define the determinant det ϕ
of ϕ to be det bϕ in F × /F ×2 if the bilinear form bϕ is non-degenerate. If char F = 2 and
dim ϕ is odd (the only remaining case), define det ϕ to be aF ×2 in F × /F ×2 where a ∈ F ×
satisfies ϕ|rad bϕ ' hai.
15. CHAIN p-EQUIVALENCE OF QUADRATIC PFISTER FORMS 59
Remark 13.9. Let ϕ be a non-degenerate even dimensional quadratic form with trivial
discriminant over F , i.e., ϕ ∈ Iq2 (F ). Then Z(ϕ) ' F × F , in particular, C(ϕ) is not a
¡ ¢
division algebra, i.e., C(ϕ) ' M2 C + (ϕ) for a central simple F -algebra C + (ϕ) uniquely
determined up to isomorphism. Moreover, C0 (ϕ) ' C + (ϕ) × C + (ϕ).
14. The Clifford invariant
A more delicate invariant of a non-degenerate even dimensional quadratic form arises

from its associated Clifford algebra.
Let ϕ be a non-degenerate even dimensional quadratic form over F . The Clifford
algebra C(ϕ) is then a central simple F -algebra. Denote by clif(ϕ) the class of C(ϕ) in
the Brauer group Br(F ). It follows from Example 11.3 that clif(ϕ) ∈ Br2 (F ). We call
clif(ϕ) the Clifford invariant of ϕ.
Example 14.1. Let ϕ be the reduced norm form of a quaternion algebra Q. It follows
from Proposition 12.4 that clif(ϕ) = [Q].
Lemma 14.2. Let ϕ and ψ be two non-degenerate even dimensional quadratic forms
over F . If disc(ϕ) is trivial then clif(ϕ ⊥ ψ) = clif(ϕ) · clif(ψ).
Proof. Let e ∈ Z(ϕ) be a nontrivial idempotent and set s = e − ē = 1 − 2e. We have
s̄ = −s, s2 = 1, and vs = s̄v = −sv for every v ∈ Vϕ by Lemma 11.7. Therefore, in
the Clifford algebra of ϕ ⊥ ψ, we have (v ⊗ 1 + s ⊗ w)2 = ϕ(v) + ψ(w) for all v ∈ Vϕ and
w ∈ Vψ . It follows from the universal property of the Clifford algebra that the F -linear
map Vϕ ⊕ Vψ → C(ϕ) ⊗F C(ψ) defined by v ⊕ w 7→ v ⊗ 1 + s ⊗ w extends to an F -algebra
homomorphism C(ϕ ⊥ ψ) → C(ϕ) ⊗F C(ψ). This map is an isomorphism as the Clifford
algebra of an even dimensional form is central simple. ¤
Theorem 14.3. The map
e2 : Iq2 (F ) → Br2 (F )
taking a form ϕ to clif(ϕ) is a well-defined group homomorphism. Moreover, Iq3 (F ) ⊂
ker(e2 ).
Proof. It follows from Lemma 14.2 that e2 is well-defined. Next let ϕ ∈ Iq2 (F ) and
a ∈ F × . Since
¡ disc(ϕ)¢ is trivial, it follows from Corollary 11.8 that C(aϕ) ' C(ϕ).
Therefore, e2 hhaii ⊗ ϕ = e2 (ϕ) − e2 (aϕ) = 0. ¤
We shall, in fact, prove that Iq3 (F ) = ker(e2 ) in §16 for fields of characteristic two and
Chapter VIII for fields of characteristic not two.
15. Chain p-equivalence of quadratic Pfister forms
We saw that anisotropic bilinear Pfister forms hha1 , . . . , an ii and hhb1 , . . . , bn ii were
p-chain equivalent if and only if they were isometric. This equivalence relation was based
on isometries of 2-fold Pfister forms. In this section, we prove the analogous result for
quadratic Pfister forms. This was first proven in the case of characteristic two by Aravire
and Baeza in [15]. To begin we therefore need to establish isometries of quadratic 2-fold
Pfister forms in characteristic two. This is given by the following:
Lemma 15.1. Let F be a field of characteristic 2. Then in Iq (F ), we have

(1) hha, b + b0 ]] = hha, b]] + hha, b0 ]].
(2) hhaa0 , b]] ≡ hha, b]] + hha0 , b]] mod Iq3 (F ).
ab
(3) hha + x2 , b]] = hha, ]].
a + x2
ab 0 a0 b
(4) hha + a0 , b]] ≡ hha, ]] + hha , ]] mod Iq3 (F ).
a + a0 a + a0
Proof. (1): This follows by Example 7.24.
(2): Follows from the equality hhaii+hha0 ii = hhaa0 ii+hha, a0 ii in the Witt ring of bilinear
forms by Example 4.10.
(3): Let c = b/(a + x2 ) and
h a, c i · ¸
a + x2 , c
A= and B= .
F F
By Corollary 12.5, it suffices to prove that A ' B. Let {1, i, j, ij} be the standard basis
of A, i.e., i2 = a, j 2 = b and ij + ji = 1. Considering the new basis {1, i + x, j, (i + x)j}
with (i + x)2 = a + x2 shows that A ' B.
(4): We have by (1)–(3):
a a ab
hha + a0 , b]] ≡ hh 0
+ 1, b]] + hha0 , b]] = hh 0 , ]] + hha0 , b]]
a a a + a0
ab 0 ab
≡ hha, ]] + hha , ]] + hha0 , b]]
a + a0 a + a0
ab 0 a0 b
= hha, ]] + hha , ]] mod Iq3 (F ). ¤
a + a0 a + a0
The definition of chain p-equivalence for quadratic Pfister forms is slightly more in-
volved then that for bilinear Pfister forms.
Definition 15.2. Let a1 , . . . , an−1 , b1 , . . . , bn−1 ∈ F × and an , bn ∈ F with n ≥ 1.
We assume that an and bn are nonzero if char F 6= 2. Let ϕ = hha1 , . . . , an−1 , an ]] and
ψ = hhb1 , . . . , bn−1 , bn ]]. We say that the quadratic Pfister forms ϕ and ψ are simply
p-equivalent if either n = 1 and hha1 ]] ' hhb1 ]] or n ≥ 2 and there exist i and j with
1 ≤ i < j ≤ n satisfying
(15.2a) hhai , aj ii ' hhbi , bj ii with j < n and al = bl for all l 6= i, j or
(15.2b) hhai , aj ]] ' hhbi , bj ]] with j = n and al = bl for all l 6= i, j.
We say that ϕ and ψ are chain p-equivalent if there exist quadratic n-fold Pfister forms
ϕ0 , . . . , ϕm for some m such that ϕ = ϕ0 , ψ = ϕm and ϕi is simply p-equivalent to ϕi+1
for each i = 0, . . . , m − 1.
Theorem 15.3. Let ϕ = hha1 , . . . , an−1 , an ]] and ψ = hhb1 , . . . , bn−1 , bn ]] be anisotropic
quadratic n-fold Pfister forms. Then ϕ ≈ ψ if and only if ϕ ' ψ.
15. CHAIN p-EQUIVALENCE OF QUADRATIC PFISTER FORMS 61
We shall prove this result in a series of steps. Suppose that ϕ ' ψ. The case char F 6= 2
was considered in Theorem 6.10, so we may also assume that char F = 2. As before the
map ℘ : F → F is defined by ℘(x) = x2 + x when char F = 2.
Lemma 15.4. Let char F = 2. If b = hha1 , . . . , an ii is an anisotropic bilinear Pfister
form and d1 , d2 ∈ F then b ⊗ hhd1 ]] ' b ⊗ hhd2 ]] if and only if b ⊗ hhd1 ]] ≈ b ⊗ hhd2 ]].
Proof. Assume that b ⊗ hhd1 ]] ' b ⊗ hhd2 ]]. Then the form
b ⊗ hhd1 + d2 ]] ∼ b ⊗ hhd1 ]] ⊥ b ⊗ hhd2 ]]
and is hyperbolic. By Lemma 9.12, we have d1 +d2 = x+y where x ∈ im(℘) and y ∈ D(b e 0 ).
If y = 0 then hhd1 ]] ' hhd2 ]] and we are done. So suppose that y 6= 0. By Lemma 6.11,
there is a bilinear Pfister form c such that b ≈ c ⊗ hhyii. As hhy, d1 ]] ' hhy, d2 ]], we have
b ⊗ hhd1 ]] ≈ c ⊗ hhy, d1 ]] ≈ c ⊗ hhy, d2 ]] ≈ b ⊗ hhd2 ]]. ¤
Lemma 15.5. Let char F = 2. If ρ = hhb, b2 , . . . , bn , d]] is a quadratic Pfister form then
for every a ∈ F × and z ∈ D(ρ), we have hhaii ⊗ ρ ≈ hhazii ⊗ ρ.
Proof. We proceed by induction on dim ρ. Let η = hhb2 , . . . bn , d]]. We have z =
e
x + by with x, y ∈ D(η). If y = 0 then x = z 6= 0 and by the induction hypothesis
hhaii ⊗ η ≈ hhazii ⊗ η, hence
hhaii ⊗ ρ = hha, bii ⊗ η ≈ hhaz, bii ⊗ η ≈ hhazii ⊗ ρ.
If x = 0 then z = by and by the induction hypothesis hhaii ⊗ η ≈ hhayii ⊗ η, hence
hhaii ⊗ ρ = hha, bii ⊗ η ≈ hhay, bii ⊗ η ≈ hhaz, bii ⊗ η ≈ hhazii ⊗ ρ.
Now suppose that both x and y are nonzero. As η is round, xy ∈ D(η). By the induction
hypothesis and Lemma 4.15,
hhaii ⊗ ρ = hha, bii ⊗ η ≈ hha, abii ⊗ η ≈ hhax, abyii ⊗ η
≈ hhaz, bxyii ⊗ η ≈ hhaz, bii ⊗ η = hhazii ⊗ ρ. ¤
Lemma 15.6. Let char F = 2. Let b = hha1 , . . . , am ii be a bilinear Pfister form,
ρ = hhb1 , . . . , bn , d]] a quadratic Pfister form with n ≥ 0 (if n = 0 then ρ = hhd]]), and
c ∈ F × . Suppose there exists an x ∈ D(b) with c + x 6= 0 satisfying b ⊗ hhc + xii ⊗ ρ is
anisotropic. Then
b ⊗ hhc + xii ⊗ ρ ≈ b ⊗ hhcii ⊗ ψ
for some quadratic Pfister form ψ.
Proof. We proceed by induction on the dimension of b. Suppose b = h1i. Then
x = y 2 for some y ∈ F . It follows from Lemma 15.1 that hhc + y 2 , d]] ' hhc, cd/(c + y 2 )]]
hence hhc + y 2 ii ⊗ ρ ≈ hhcii ⊗ hhb1 , . . . , bn , cd/(c + y 2 )]].
So we may assume that dim b > 1. Let c = hha2 , . . . , am ii and a = a1 . We have
e
x = y + az where y, z ∈ D(c). If c = az then c + x = y belongs to D(b), so the form
b ⊗ hhc + xii would be metabolic contradicting hypothesis.
Let d := c + az. We have d 6= 0. By the induction hypothesis,
c ⊗ hhd + yii ⊗ ρ ≈ c ⊗ hhdii ⊗ µ and c ⊗ hhac + a2 zii ⊗ µ ≈ c ⊗ hhacii ⊗ ψ
for some quadratic Pfister forms µ and ψ. Hence by Lemma 4.15,

b ⊗ hhc + xii ⊗ ρ = b ⊗ hhd + yii ⊗ ρ = c ⊗ hha, d + yii ⊗ ρ
≈ c ⊗ hha, dii ⊗ µ = c ⊗ hha, c + azii ⊗ µ ≈ c ⊗ hha, ac + a2 zii ⊗ µ
≈ c ⊗ hha, acii ⊗ ψ ≈ c ⊗ hha, cii ⊗ ψ = b ⊗ hhcii ⊗ ψ. ¤
If b is a bilinear Pfister form over a field F then b ' b0 ⊥ h1i with the pure subform
b0 unique up to isometry. For quadratic Pfister form over a field of characteristic two,
the analogue of this is not true. So, in this case, we have to modify our notion of a pure
subform of a quadratic Pfister form. So suppose that char F = 2. Let ϕ = b ⊗ hhd]] be a
quadratic Pfister form with b = b0 ⊥ h1i, a bilinear Pfister form. We have ϕ = hhd]] ⊥ ϕ◦
with ϕ◦ = b0 ⊗ hhd]]. The form ϕ◦ depends on the presentation of b. Let ϕ0 := h1i ⊥
b0 ⊗ hhd]]. This form coincides with the complementary form h1i⊥ in ϕ. The form ϕ0 is
uniquely determined by ϕ up to isometry. Indeed, by Witt Extension Theorem 8.3, for
any two vectors v, w ∈ Vϕ with ϕ(v) = ϕ(w) = 1 there is an auto-isometry α of ϕ such
that α(v) = w. Therefore, the orthogonal complements of F v and F w are isometric. We
call the form ϕ0 the pure subform of ϕ.
Proposition 15.7. Let ρ = hhb1 , . . . , bn , d]] be a quadratic Pfister form, n ≥ 1, and
b = hha1 , . . . , am ii a bilinear Pfister form. Set ϕ = b ⊗ ρ. Suppose that ϕ is anisotropic.
Let c ∈ D(b ⊗ ρ0 ) \ D(b). Then ϕ ≈ b ⊗ hhcii ⊗ ψ for some quadratic Pfister form ψ.
Proof. Let ρ = hhbii ⊗ η with b = b1 and η = hhb2 , . . . , bn , d]] (with η = hhd]] if
n = 1). We proceed by induction on dim ρ. Note that if n = 1, we have η 0 = h1i and
D(b ⊗ η 0 ) = D(b). As
b ⊗ ρ0 = (b ⊗ η 0 ) ⊥ (bb ⊗ η),
e ⊗ η 0 ), and y ∈ D(b
we have c = x + by with x ∈ D(b e ⊗ η).
If y = 0 then c = x ∈ D(b ⊗ η 0 ) \ D(b). In particular, n > 1. By the induction
hypothesis, b ⊗ η ≈ b ⊗ hhcii ⊗ µ for some quadratic Pfister form µ. Hence
ϕ = b ⊗ ρ = b ⊗ hhbii ⊗ η ≈ b ⊗ hhcii ⊗ hhbii ⊗ µ.
Now suppose that y 6= 0. By Lemma 15.5,
(15.8) ϕ = b ⊗ ρ = b ⊗ hhbii ⊗ η ≈ b ⊗ hhbyii ⊗ η.
Assume that x ∈ e
/ D(b). In particular, n > 1. By the induction hypothesis, b ⊗ η ≈
b ⊗ hhxii ⊗ µ for some quadratic Pfister form µ. Therefore by Lemma 4.15,
ϕ ≈ b ⊗ hhbyii ⊗ η ≈ b ⊗ hhx, byii ⊗ µ ≈ b ⊗ hhc, bxyii ⊗ µ.
e
Finally we assume that x ∈ D(b). By Lemma 15.6 and (15.8),
ϕ ≈ b ⊗ hhbyii ⊗ η = b ⊗ hhc + xii ⊗ η ≈ b ⊗ hhcii ⊗ ψ
for a quadratic Pfister form ψ. ¤
Proof. (of Theorem 15.3) Let ϕ and ψ be isometric anisotropic quadratic n-fold
Pfister forms over F as in the statement of Theorem 15.3. We must show that ϕ ≈ ψ.
We may assume that char F = 2.
16. COHOMOLOGICAL INVARIANTS 63
Claim: For every r = 0, . . . , n − 1 there exist a bilinear r-fold Pfister form b and quadratic
(n − r)-fold Pfister forms ρ and µ such that ϕ ≈ b ⊗ ρ and ψ ≈ b ⊗ µ.
We prove the claim by induction on r. The case r = 0 is obvious. Suppose we have such
b, ρ and µ for some r < n − 1. Write ρ = hhcii ⊗ ψ for some c ∈ F × and quadratic Pfister
form ψ, so ϕ ≈ b ⊗ hhcii ⊗ ψ. Note that as ϕ is anisotropic, we have c ∈ D(b ⊗ ρ0 ) \ D(b).
The form b⊗h1i is isometric to subforms of ϕ and ψ. As bϕ and bψ are non-degenerate,
by the Witt Extension Theorem 8.3, an isometry between these subforms extends to an
isometry between ϕ and ψ. This isometry induces an isometry of orthogonal complements
b ⊗ ρ0 and b ⊗ µ0 . Therefore, we have c ∈ D(b ⊗ ρ0 ) \ D(b) = D(b ⊗ µ0 ) \ D(b). It follows
from Proposition 15.7 that ψ ≈ b ⊗ hhcii ⊗ σ for some quadratic Pfister form σ. Thus
ψ ≈ b ⊗ hhcii ⊗ σ and the claim is established.
Applying the claim in the case r = n − 1, we find a bilinear (n − 1)-fold Pfister form
b and elements d1 , d2 ∈ F such that ϕ ≈ b ⊗ hhd1 ]] and ψ ≈ b ⊗ hhd2 ]]. By Lemma 15.4,
we have b ⊗ hhd1 ]] ≈ b ⊗ hhd2 ]], hence ϕ ≈ ψ. ¤
16. Cohomological invariants
A major problem in the theory of quadratic forms was to determine the relationship
between quadratic forms and Galois cohomology. In this section, using the cohomology
groups defined in §101, we introduce the problem.
Let H ∗ (F ) be the groups defined in §101. In particular,
½
n Ét2 (F ), if n = 1.
H (F ) '
Br2 (F ), if n = 2.
Let ϕ be a quadratic n-fold Pfister form. Suppose that ϕ ' hha1 , . . . , an ]]. Define the
cohomological invariant of ϕ to be the class en (ϕ) in H n (F ) given by
en (ϕ) = {a1 , a2 , . . . , an−1 } · [Fan ],
where [Fc ] is the class of the étale quadratic extension Fc /F in Ét2 (F ) ' H 1 (F ).
The cohomological invariant en is well-defined on quadratic n-fold Pfister forms:
Proposition 16.1. Let ϕ and ψ be quadratic n-fold Pfister forms. If ϕ ' ψ then
en (ϕ) = en (ψ) in H n (F ).
Proof. This follows from Theorems 6.20 and 15.3. ¤

As in the bilinear case, if we use the Hauptsatz 23.8 below, we even have if
ϕ ≡ ψ mod Iqn+1 (F ) then en (ϕ) = en (ψ)
in H n (F ). (Cf. Corollary 23.10 below.) In fact, we shall also show by elementary means
in Proposition 24.6 below that if ϕ1 , ϕ2 and ϕ3 are general quadratic n-fold Pfister forms
such that ϕ1 + ϕ2 + ϕ3 ∈ Iqn+1 (F ) then en (ϕ1 ) + en (ϕ2 ) + en (ϕ3 ) = 0 ∈ H n (F ).
We call the extension of en to a group homomorphism en : Iqn (F ) → H n (F ) the nth
cohomological invariant of Iqn (F ).
Fact 16.2. The nth cohomological invariant en exists for all fields F and for all n ≥ 1.
Moreover, ker(en ) = Iqn+1 (F ). Further, there is a unique isomorphism
en : Iqn (F )/Iqn+1 (F ) → H n (F )
¡ ¢
satisfying ēn ϕ + Iqn+1 (F ) = en (ϕ) for every quadratic n-fold Pfister quadratic form ϕ.
Special cases of Fact 16.2 can be proven by elementary methods. Indeed we have
already shown that the invariant e1 is well-defined on all of Iq (F ) and coincides with the
discriminant in Theorem 13.7 and e2 is well-defined on all of Iq2 (F ) and coincides with
the Clifford invariant by Theorem 14.3. Then by Theorems 13.7 and 14.3 the maps ē1
and ē2 are well-defined. For fields of characteristic different from two, e3 was shown to be
well-defined hence ē3 by Arason in [6] and ē3 an isomorphism in [105] and independently
by Rost [120]; and Jacob and Rost showed that e4 was well defined in [67].
Suppose that char F 6= 2. Then the identification of bilinear and quadratic forms leads
to the composition
fn e
hnF : Kn (F )/2Kn (F ) −→ I n (F )/I n+1 (F ) = Iqn (F )/Iqn+1 (F ) −→
n
H n (F ).
where hnF is the norm residue homomorphism of degree n defined in §101.
Milnor conjectured that hnF was an isomorphism for all n in [108]. Voevodsky proved
the Milnor Conjecture in [142]. As was stated in Fact 5.15 the map fn is an isomorphism
for all n. In particular, en is well-defined and ēn is an isomorphism for all n.
If char F = 2, Kato proved Fact 16.2 in [77].
We have proven that ē1 is an isomorphism in Theorem 13.7. We shall prove that h2F
is an isomorphism in Chapter VIII below if the characteristic of F is different from two.
It follows that ē2 is an isomorphism. We now turn to the proof that ē2 is an isomorphism
if char F = 2. The following statement was first proven by Sah in [122].
Theorem 16.3. Let char F = 2. Then ē2 : Iq2 (F )/Iq3 (F ) → Br2 (F ) is an isomorphism.
Proof. The classes of quaternion algebras generate the group Br2 (F ) by [1, Ch. VII,
Th. 30]. It follows that ē2 is surjective. So we need only show that ē2 is injective.
P
Let α ∈ Iq2 (F ) satisfy e2 (α) = 1. Write α in the form ni=1 di hhai , bi ]]. By assumption,
ha , c i
i i
the product of all with ci = bi /ai is trivial in Br(F ).
F
3
We prove i induction on n that α ∈ Iq (F ). If n = 1, we have α = hha1 , b1 ]] and
h a , c by
1 1
e2 (α) = = 1. Therefore, the reduced norm form α of the split quaternion algebra
ha , c i F
1 1
is hyperbolic by Corollary 12.5, hence α = 0.
F
ha , c i
1/2 1/2 i i
In the general case, let L = F (a1 , . . . , an−1 ). The field L splits for all
ha , c i h a , c Fi · c, d ¸
n n n n
i = 1, . . . , n − 1 and hence splits . By Lemma 98.17, we have = ,
F F F
where c is the square of an element of L, i.e., c is the sum of elements of the form g 2 m
16. COHOMOLOGICAL INVARIANTS 65
where g ∈ F and m is a monomial in the ai , i = 1, . . . , n−1. It follows from Corollary 12.5

that hhan , bn ]] = hhc, cd]]. By Lemma 15.1, we have hhc, cd]] is congruent modulo Iq3 (F ) to
the sum of 2-fold P Pfister forms hhai , fi ]] with i = 1, . . . , n − 1, fi ∈ F . Therefore we may
assume that α = n−1 0 0 3
i=1 hhai , bi ]] for some bi . By the induction hypothesis, α ∈ Iq (F ). ¤
CHAPTER III
Forms over Rational Function Fields
17. The Cassels-Pfister Theorem

Given a quadratic form ϕ over a field over F , it is natural to consider values of the form
over the rational function field F (t). The Cassels-Pfister Theorem shows that whenever
ϕ represents a polynomial over F (t) then it already does so when viewed as a quadratic
form over the polynomial ring F [t]. This results in specialization theorems. As an n-
dimensional quadratic form ψ can be viewed as a polynomial in F [T ] := F [t1 , . . . , tn ],
one can also ask when is ψ(T ) a value of ϕF (T ) ? If both the forms are anisotropic, we
shall also show in this section the fundamental result that this is true if and only if ψ is
a subform of ϕ.
Computation 17.1. Let ϕ be an anisotropic quadratic form on V over F and b its
polar form. Let v and u be two distinct vectors in V and set w = v − u. Let τw be the
reflection with respect to w as defined in Example 7.3. Then
¡ ¢
(1) ϕ τw (v) = ϕ(v) as τw is an isometry.
ϕ(u) − ϕ(v)
(2) τw (v) = u + w as bϕ (v, w) = −bϕ (v, −w) = −ϕ(u) + ϕ(v) + ϕ(w) by
ϕ(w)
definition.
Notation 17.2. If T = (t1 , . . . , tn ) is a tuple of independent variables, let
F [T ] := F [t1 , . . . , tn ] and F (T ) := F (t1 , . . . , tn ).
If V is a finite dimensional vector space over F , let

V [T ] := F [T ] ⊗F V and V (T ) := VF (T ) := F (T ) ⊗F V.
Note that V (T ) is also the localization of V [T ] at F [T ] \ {0}. In particular, if v ∈ V (T )
there exist w ∈ V [T ] and a nonzero f ∈ F [T ] satisfying v = w/f . For a single variable t,
we let V [t] := F [t] ⊗F V and V (t) := VF (t) := F (t) ⊗F V .
The following general form of the Classical Cassels-Pfister Theorem is true.
Theorem 17.3. (Cassels-Pfister Theorem) Let ϕ be a quadratic form on V and let
h ∈ F [t] ∩ D(ϕF (t) ). Then there is a w ∈ V [t] such that ϕ(w) = h.
Proof. Suppose first that ϕ is anisotropic. Let v ∈ V (t) satisfy ϕ(v) = h. There is
a nonzero polynomial f ∈ F [t] such that f v ∈ V [t]. Choose v and f so that deg(f ) is the
smallest possible. It suffice to show that f is constant. Suppose deg(f ) > 0.
67
68 III. FORMS OVER RATIONAL FUNCTION FIELDS
Using the analog of the Division Algorithm, we can divide the polynomial vector
f v by f to get f v = f u + r, where u, r ∈ V [t] and deg(r) < deg(f ). If r = 0 then
v = u ∈ V [t] and f is constant; so we may assume that r 6= 0. In particular, ϕ(r) 6= 0 as
ϕ is anisotropic. Set w = v − u = r/f and consider
ϕ(u) − h
(17.4) τw (v) = u + r
ϕ(r)/f
¡ ¢
as in Computation 17.1(2). We have ϕ τw (v) = h. We show that f 0 := ϕ(r)/f is a
polynomial. As
f 2 h = ϕ(f v) = ϕ(f u + r) = f 2 ϕ(u) + f bϕ (u, r) + ϕ(r),
we see that ϕ(r) is divisible by f . Equation (17.4) implies that f 0 τw (v) ∈ V [t] and the
definition of r yields
deg(f 0 ) = deg ϕ(r) − deg(f ) < 2 deg(f ) − deg(f ) = deg(f ),
a contradiction to the minimality of deg(f ).
Now suppose that ϕ is isotropic. By Lemma 7.13, we may assume that rad ϕ = 0. In
particular, a hyperbolic plane splits off as an orthogonal direct summand of ϕ by Lemma
7.14. Let e, e0 be a hyperbolic pair for this hyperbolic plane. Then
ϕ(he + e0 ) = bϕ (he, e0 ) = hbϕ (e, e0 ) = h. ¤
The theorem above was first proved by Cassels in [24] for the form h1, . . . , 1i over a
field of characteristic not two. This was generalized by Pfister to non-degenerate forms
over such fields in [111] and used to prove the the results through Corollary 17.13 below
in that case.
Corollary 17.5. Let b be a symmetric bilinear form on V and let h ∈ F [t] ∩
D(ϕF (t) ). Then there is a v ∈ V [t] such that b(v, v) = h.
Proof. Let ϕ be ϕb , i.e., ϕ(v) = b(v, v) for all v ∈ V . As D(ϕ) = D(b) by Lemma
9.3, the result follows from the Cassels-Pfister Theorem. ¤
Corollary 17.6. Let f ∈ F [t] be a sum of n squares in F (t). Then f is a sum of n
squares in F [t].
Corollary 17.7. (Substitution Principle) Let ϕ be a quadratic form over F and
h ∈ D(ϕF (T ) ) with T = (t1 , . . . , tn ). Suppose that h(x) is defined for x ∈ F n and h(x) 6= 0
then h(x) ∈ D(ϕ).
Proof. As h(x) is defined, we can write h = f /g with f, g ∈ F [T ] and g(x) 6= 0.
Replacing h by g 2 h, we may assume that h ∈ F [T ]. Let T 0 = (t1 , . . . , tn−1 ¡ ) and x¢ =
0 0 0
(x1 , . . . , xn ). By the theorem, there exists v(T , tn ) ∈ V (T¡)[tn ] satisfying
¢ ϕ v(T , tn ) =
0 0 0
h(T , tn ). Evaluating tn at xn shows that h(T , xn ) = ϕ v(T , xn ) ∈ D(ϕF (T 0 ) ). The
conclusion follows by induction on n. ¤
As above, we also deduce:
Corollary 17.8. (Bilinear Substitution Principle) Let b be a symmetric bilinear form
over F and h ∈ D(bF (T ) ) with T = (t1 , . . . , tn ). Suppose that h(x) is defined for x ∈ F n
and h(x) 6= 0 then h(x) ∈ D(b).
17. THE CASSELS-PFISTER THEOREM 69
We shall need the following slightly more general version of the Cassels-Pfister Theo-
rem.
Proposition 17.9. Let ϕ be an anisotropic quadratic form on V . Suppose that s ∈ V
and v ∈ V (t) satisfy ϕ(v) ∈ F [t] and bϕ (s, v) ∈ F [t]. Then there is w ∈ V [t] such that
ϕ(w) = ϕ(v) and bϕ (s, w) = bϕ (s, v).
Proof. It suffices to show the value bϕ (s, v) does not change when v is modified in
the course of the proof of Theorem 17.3. Choose v0 ∈ V [t] satisfying bϕ (s, v0 ) = bϕ (s, v).
Let f ∈ F [t] be a nonzero polynomial such that f v ∈ V [t]. As the remainder r on
⊥ ⊥
dividing f v and
¡ f v − f¢v0 by f is the same and f v − f v0 ∈ (F (t)s) , we have r ∈ (F (t)s) .
Therefore, bϕ s, τr (v) = bϕ (s, v). ¤
Lemma 17.10. Let ϕ be an anisotropic quadratic form and ρ a non-degenerate binary
anisotropic quadratic form satisfying ρ(t1 , t2 ) + d ∈ D(ϕF (t1 ,t2 ) ) for some d ∈ F . Then
e
ϕ ' ρ ⊥ µ for some form µ and d ∈ D(µ).
Proof. Let ρ(t1 , t2 ) = at21 +bt1 t2 +ct22 . As ρ(t1 , t2 )+dt23 is a value of ϕ over F (t1 , t2 , t3 ),
there is a u ∈ V = Vϕ such that ϕ(u) = a by the Substitution Principle 17.7. Applying
the Cassels-Pfister Theorem 17.3 to the form ϕF (t2 ) , we find a v ∈ VF (t2 ) [t1 ] such that
ϕ(v) = at21 +bt1 t2 +ct22 +d. Since ϕ is anisotropic, we have degt1 v ≤ 1, i.e., v(t1 ) = v0 +v1 t1
for some v0 , v1 ∈ VF (t2 ) . Expanding we get
ϕ(v0 ) = a, b(v0 , v1 ) = bt2 , ϕ(v1 ) = ct22 + d,
where b = bϕ . Clearly v0 ∈
/ rad(bF (t2 ) ).
We claim that u ∈ / rad(b). We may assume that u 6= v0 and therefore
0 6= ϕ(u − v0 ) = ϕ(u) + ϕ(v0 ) − b(u, v0 ) = b(u, u − v0 )
as ϕF (t2 ) is anisotropic by Lemma 7.16 hence the claim follows.
By the Witt Extension Theorem 8.3, there is an isometry γ of ϕF (t2 ) satisfying γ(v0 ) =
u. Replacing v0 and v1 by u = γ(v0 ) and γ(v1 ) respectively, we may assume that v0 ∈ V .
Applying Proposition 17.9 to the vectors v0 and v1 , we find w ∈ V [t2 ] satisfying
ϕ(w) = ct22 + d and b(v0 , w) = bt2 . In a similar fashion, we have w = w0 + w1 t2 with
w0 , w1 ∈ V . Expanding, we have
ϕ(v0 ) = a, b(v0 , w1 ) = b, ϕ(w1 ) = c, ϕ(w0 ) = d, b(v0 , w0 ) = 0,b(w0 , w1 ) = 0.
It follows if W is the subspace generated by v0 and w1 then ϕ|W e
' ρ and d ∈ D(µ) where
µ = ϕ|W ⊥ . ¤
Corollary 17.11. Let ϕ and ψ be two anisotropic quadratic forms over F with
dim ψ = n. Let T = (t1 , . . . , tn ). Suppose that ψ(T ) ∈ D(ϕF (T ) ). If ψ = ρ ⊥ σ with ρ a
non-degenerate binary form and T 0 = (t3 , . . . , tn ) then ϕ ' ρ ⊥ µ for some form µ and
e F (T 0 ) (ϕF (T 0 ) ).
µ(T 0 ) ∈ D
Theorem 17.12. (Representation Theorem) Let ϕ and ψ be two anisotropic quadratic
forms over F with dim ψ = n. Let T = (t1 , . . . , tn ). Then the following are equivalent:
(1) D(ψK ) ⊂ D(ϕK ) for every field extension K/F .
(2) ψ(T ) ∈ D(ϕF (T ) ).

(3) ψ is isometric to a subform of ϕ.
In particular, if any of the above conditions hold then dim ψ ≤ dim ϕ.
Proof. (1) ⇒ (2) and (3) ⇒ (1) are trivial.
(2) ⇒ (3): Applying the structure results, Propositions 7.32 and 7.30, we can write
ψ = ψ1 ⊥ ψ2 , where ψ1 is an orthogonal sum of non-degenerate binary forms and ψ2 is
diagonalizable. Repeated application of Corollary 17.11 allows us to reduce to the case
ψ = ψ2 , i.e., ψ = ha1 , . . . , an i is diagonalizable.
We proceed by induction on n. The case n = 1 follows from the Substitution Principle
17.7. Suppose that n = 2. Then we have a1 t2 + a2 ∈ D(ϕF (t) ). By the Cassels-Pfister
Theorem, there is a v ∈ V [t] where V = Vϕ satisfying ϕ(v) = a1 t2 +a2 . As ϕ is anisotropic,
we have v = v1 +v2 t for v1 , v2 ∈ V and therefore ϕ(v1 ) = a1 , ϕ(v2 ) = a2 , and b(v1 , v2 ) = 0.
The restriction of ϕ on the subspace spanned by v1 and v2 is isometric to ψ.
In the general case, set T = (t1 , t2 , . . . , tn ), T 0 = (t2 , . . . , tn ), b = a2 t2 + · · · + an t2n .
As a1 t2 + b is a value of ϕ over F (T 0 )(t), by the case considered above, there are vectors
v1 , v2 ∈ VF (T 0 ) satisfying
ϕ(v1 ) = a1 , ϕ(v2 ) = b and b(v1 , v2 ) = 0.
It follows from the Substitution Principle 17.7 that there is w ∈ V such that ϕ(w) = a1 .
We claim that there is an isometry γ of ϕ over F (T 0 ) such that ϕ(v1 ) = w. We may
assume that w 6= v1 as ϕF (T 0 ) is anisotropic by Lemma 7.16. We have
0 6= ϕ(w − v1 ) = ϕ(w) + ϕ(v1 ) − b(w, v1 ) = b(w, w − v1 ) = b(v1 − w, v1 ),
therefore w and v1 do not belong to rad b. The claim follows by the Witt Extension
Theorem 8.3.
Replacing v1 and v2 by γ(v1 ) = w and γ(v2 ) respectively, we may assume that v1 ∈ V .
Set W = (F v1 )⊥ . Note that v2 ∈ WF (T 0 ) , hence b is a value of ϕ|W over F (T 0 ). By the
induction hypothesis applied to the forms ψ 0 = ha2 , . . . , an i and ϕ|W , there is a subspace
V 0 ⊂ W such that ϕ|V 0 ' ha2 , . . . , an i. Note that v1 is orthogonal to V 0 and v1 ∈ / V 0 as
ψ is anisotropic. Therefore, the restriction of ϕ on the subspace F v1 ⊕ V 0 is isometric to
ψ. ¤
A field F is called formally real if −1 is not a sum of squares. In particular, char F = 0
if this is the case. (Cf. §95.)
Corollary 17.13. Suppose that F is formally real and T = (t0 , . . . , tn ). Then
t20 + t21 + · · · + t2n is not a sum of n squares in F (T ).
¡ ¢
Proof. If this is false then t20 + t21 + · · · + t2n ∈ D nh1i . As (n + 1)h1i is anisotropic,
this contradicts the Representation Theorem. ¤
The ideas above also allow us to develop a test for simultaneous zeros for quadratic
forms. This was proven independently by Amer in [2] and Brumer in [23] for fields of
characteristic not two and by Leep for arbitrary fields in [93].
18. VALUES OF FORMS 71
Theorem 17.14. Let ϕ and ψ be two quadratic forms on a vector space V over F .
Then the form ϕF (t) + tψF (t) on V (t) over F (t) is isotropic if and only if ϕ and ψ have a
common isotropic vector in V .
Proof. Clearly, a common isotropic vector for ϕ and ψ is also an isotropic vector for
ρ := ϕF (t) + tψF (t) .
Conversely, let ρ be isotropic. There exists a nonzero v ∈ V [t] such that ρ(v) = 0.
Choose such a v of the smallest degree. We claim that deg v = 0, i.e., v ∈ V . If we show
this, the equality ϕ(v) + tψ(v) = 0 implies that v is a common isotropic vector for ϕ and
ψ.
Suppose n := deg v > 0. Write v = w + tn u with u ∈ V and w ∈ V [t] of degree less
than n. Note that by assumption ρ(u) 6= 0. Consider the vector
v 0 = ρ(u) · τu (v) = ρ(u)v − bρ (v, u)u ∈ V [t].
As ρ(v) = 0, we have ρ(v 0 ) = 0. It follows from the equality
¡ ¢
ρ(w)v − bρ (v, w)w = ρ(v − tn u)v − bρ (v, v − tn u)(v − tn u) = t2n ρ(u)v − bρ (v, u)u
that
ρ(w)v − bρ (v, w)w
v0 = .
t2n
Note that deg ρ(w) ≤ 2n − 1 and deg bρ (v, w) ≤ 2n. Therefore deg v 0 < n, contradicting
the minimality of n. ¤
18. Values of forms
Let ϕ be an anisotropic quadratic form over F . Let p ∈ F [T ] = F [t1 , . . . , tn ] be

irreducible and F (p) the quotient field of F [T ]/(p). In this section, we determine what
it means for ϕF (p) to be isotropic. We base our presentation on ideas of Knebusch in
[82]. This result has consequences for finite extensions K/F . In particular, the classical
Springer’s Theorem that forms remain anisotropic under odd degree extensions follows as
well as a norm principle about values of ϕK .
Order the group Zn lexicographically, i.e., (i1 , . . . , in ) < (j1 , . . . , jn ) if for the first
integer k satisfying ik 6= jk with 1 ≤ k ≤ n we have ik < jk . If i = (i1 , . . . , in ) in
Zn and a ∈ F × , write aT i for ati11 · · · tinn and call i the degree of aT i . Let f = aT i +
monomials of lower degree in F [T ] with a ∈ F × . The term aT i is called the leading
term of f . We define the degree deg f of f to be i, the degree of the leading term,
and the leading coefficient f ∗ of f to be a, the coefficient of the leading term. Let
Tf denote T i if i is the degree of the leading term of f . Then f = f ∗ Tf + f 0 with
deg f 0 < deg Tf . For convenience, we view deg 0 < deg f for every nonzero f ∈ F [T ].
Note that deg(f g) = deg f + deg g and (f g)∗ = f ∗ g ∗ . If h ∈ F (T )× and h = f /g with
f, g ∈ F [T ], let h∗ = f ∗ /g ∗ .
Let V be a finite dimensional vector space over F . For every nonzero v ∈ V [T ] define
the degree deg v, the leading vector v ∗ , and the leading term v ∗ Tv in a similar fashion. Let
deg 0 < deg v for any nonzero v ∈ V [T ]. So if v ∈ V [T ] is nonzero, we have v = v ∗ Tv + v 0
with deg v 0 < deg Tv .
Lemma 18.1. Let ϕ be a quadratic form on V over F and g ∈ F [T ]. Suppose that

g ∈ D(ϕF (T ) ). Then g ∗ ∈ D(ϕ). If, in addition, ϕ is anisotropic then deg g ∈ 2Zn .
Proof. Since ϕ on V and the induced quadratic form ϕ̄ on V / rad ϕ have the same
values, we may assume that rad(ϕ) = 0. In particular, if ϕ is isotropic it is universal, so
we may assume that ϕ anisotropic.
Let g = ϕ(v) with v ∈ V (T ). Write v = w/f with w ∈ V [T ] and nonzero f ∈ F [T ].
Then f 2 g = ϕ(w). As (f 2 g)∗ = (f ∗ )2 g ∗ , we may assume that v ∈ F [T ]. Let v = v ∗ Tv + v 0
with deg v 0 < deg v. Then
g = ϕ(v ∗ Tv ) + bϕ (v ∗ Tv , v 0 ) + ϕ(v 0 ) = ϕ(v ∗ )Tv2 + bϕ (v ∗ , v 0 )Tv + ϕ(v 0 )
= ϕ(v ∗ )Tv2 + terms of lower degree.
As ϕ is anisotropic, we must have ϕ(v ∗ ) 6= 0, hence g ∗ = ϕ(v ∗ ) ∈ D(ϕ). As the leading
term of g is ϕ(v ∗ )Tv2 , the second statement also follows. ¤
Let v ∈ V [T ]. Suppose that f ∈ F [T ] satisfies degt1 f > 0. Let T 0 = (t2 , . . . , tn ).

Viewing v ∈ V (T 0 )[t1 ], the analog of the usual division algorithm produces an equation
v = f w0 + r0 with w0 , r0 ∈ VF (T 0 ) [t1 ] and degt1 r0 < degt1 f.
Clearing denominators in F [T 0 ], we get
hv = f w + r
(18.2) with w, r ∈ V [T ], 0 6=h ∈ F [T 0 ] and degt1 r < degt1 f
so deg h < deg f, deg r < deg f.
®
If ϕ is a quadratic form over F let D(ϕ) denote the subgroup in F × generated by
D(ϕ).
Theorem 18.3. (Quadratic Value Theorem) Let ϕ be an anisotropic quadratic form
on V and f ∈ F [T ] a nonzero polynomial. Then the following conditions are equivalent:
®
(1) f ∗ f ∈ D(ϕF (T ) ) .
®
(2) There exists an a ∈ F × such that af ∈ D(ϕF (T ) ) .
(3) ϕF (p) is isotropic for each irreducible divisor p occurring to an odd power in the
factorization of f .
®
(2) ⇒ (3): Let afQ∈ D(ϕF (T ) ) , i.e., there are 0 6= h ∈ F [T ] and v1 , . . . , vm ∈ V [T ]
such that ah2 f = ϕ(vi ). Let p be an irreducible divisor of f to an odd power. Write P
vi = pki vi0 so that vi0 is notQdivisible by p. Dividing out both sides by p2k , with k = ki ,
we see that the product ϕ(vi0 ) is divisible by p. Hence the residue of one of the ϕ(vi0 )
is trivial in the residue field F (p) while the residue of vi0 is not trivial. Therefore, fF (p) is
isotropic.
(3) ⇒ (1): We proceed by induction on n and deg f . The statement is obvious if f = f ∗ .
In the general case, we may assume that f is irreducible. Therefore, by assumption ϕF (f ) is
isotropic. In particular, we see that there exists a vector v ∈ Vϕ [T ] such that f | ϕ(v) and
f 6 | v. If degt1 f = 0, let T 0 = (t2 , . . . , tn ) and let L denote the quotient field of F [T 0 ]/(f ).
18. VALUES OF FORMS 73
Then F (f ) = L(t1 ) so ϕL is isotropic by Lemma 7.16 and we are done by induction on

n. Therefore, we may assume that degt1 f > 0. By (18.2), there exist 0 6= h ∈ F [T ] and
w, r ∈ V [T ] such that hv = f w + r with deg h < deg f and deg r < deg f . As
ϕ(hv) = ϕ(f w + r) = f 2 ϕ(w) + f bϕ (w, r) + ϕ(r),
we have f | ϕ(r). If r = 0 then f | hv. But f is irreducible and f 6 | v so f | h. This is
impossible as deg h < deg f . Thus r 6= 0. Let ϕ(r) = f g for some g ∈ F [T ]. As ϕ is
anisotropic g 6= 0. So we have f g ∈ D(ϕF (T ) ) hence also (f g)∗ = f ∗ g ∗ ∈ D(ϕ) by Lemma
18.1.
Let p be an irreducible divisor occurring to an odd power in the factorization of g.
As deg ϕ(r) < 2 deg f , we have deg g < deg f hence p occurs with the same multiplicity
in the factorization of f g. By (2) ⇒ (3), applied to the polynomial f g, the form ϕF (p) is
isotropic. Hence the induction hypothesis implies ® that g ∗ g ∈ hD(ϕF (T ) )i. Consequently,
f ∗ f = f ∗ 2 · (f ∗ g ∗ )−1 · g ∗ g · f g · g −2 ∈ D(ϕF (T ) ) . ¤
Theorem 18.4. (Bilinear Value Theorem) Let b be an anisotropic symmetric bilin-
ear form on V and f ∈ F [T ] a nonzero polynomial. Then the following conditions are
equivalent:
®
(1) f ∗ f ∈ D(bF (T ) ) .
®
(2) There exists an a ∈ F × such that af ∈ D(bF (T ) ) .
(3) bF (p) is isotropic for each irreducible divisor p occurring to an odd power in the
factorization of f .
Proof. Let ϕ = ϕb . As D(bK ) = D(ϕK ) for every field extension K/F by Lemma
9.3 and bK is isotropic if and only if ϕK is isotropic, the result follows by the Quadratic
Value Theorem 18.3. ¤
A basic result in Artin-Schreier theory is that an ordering on a formally real field
extends to an ordering on a finite algebraic extension of odd degree, equivalently if the
bilinear form nh1i is anisotropic over F for any integer n, it remains so over any finite
extension of odd degree. Witt conjectured in [146] that any anisotropic symmetric bilinear
form remains anisotropic under a odd degree extension (if char F 6= 2). This was first
shown to be true by Springer in [132]. This is in fact true without a characteristic
assumption for both quadratic and symmetric bilinear forms.
Corollary 18.5. (Springer’s Theorem) Let K/F be a finite extension of odd degree.
Suppose that ϕ (respectively, b) is an anisotropic quadratic form (respectively, symmetric
bilinear form) over F . Then ϕK (respectively, bK ) is anisotropic.
Proof. By induction on [K : F ], we may assume that K = F (θ) is a primitive
extension. Let p be the
® minimal polynomial of θ over F . Suppose that ϕK is isotropic.
Then ap ∈ D(ϕF (t) ) for some a ∈ F × by the Quadratic Value Theorem 18.3. It follows
that p has even degree by Lemma 18.1, a contradiction. If b is a symmetric bilinear form
over F , applying the above to the quadratic form ϕb shows the theorem also holds in the
bilinear case. ¤
We shall give another proof of Springer’s Theorem in Corollary 71.3 below.
Corollary 18.6. If K/F is an extension of odd degree then rK/F : W (F ) → W (K)

and rK/F : Iq (F ) → Iq (K) are injective.
Corollary 18.7. Let ϕ and ψ be two quadratic forms on a vector space V over F
having no common isotropic vector in V . Then for any field extension K/F of odd degree
the forms ϕK and ψK have no common isotropic vector in VK .
Proof. This follows from Springer’s Theorem and Theorem 17.14. ¤
Exercise 18.8. Let char F 6= 2 and K/F be a finite purely inseparable field extension.
Then rK/F : W (F ) → W (K) is an isomorphism.
Corollary 18.9. Let K = F (θ) be an algebraic extension of F and p the (monic)
minimal polynomial of θ over F . Let ϕ be a®regular quadratic form over F . Suppose that
there exists a c ∈ F such that p(c) ∈
/ D(ϕ) . Then ϕK is anisotropic.
Proof. As rad ϕ = 0, if ϕ were isotropic it would be universal. Thus ϕ is anisotropic.
In particular, p is not linear hence p(c) 6= 0. Suppose ® that ϕK is isotropic. By the
Quadratic Value Theorem
18.3,
® we have p ∈ D(ϕ F (t) . By the Substitution Principle
)
17.7, we have p(c) ∈ D(ϕ) for all c ∈ F , a contradiction. ¤
As another consequence, we obtain the following theorem first proved by Knebusch in
[81].
Theorem 18.10. (Value Norm Principle)
¡ ¢ Let
ϕ be ®a quadratic form over F and‘ K/F
a finite field extension. Then NK/F D(ϕK ) ⊂ D(ϕ) .
Proof. Let V = Vϕ . Since the form ϕ on V and the induced form ϕ̄ on V / rad(ϕ)
have the same values, we may assume that rad(ϕ) = 0. If ϕ is isotropic then ϕ splits off a
hyperbolic plane. In particular, ϕ is universal and the statement is obvious. Thus we may
assume that ϕ is anisotropic. Moreover, we may assume that dim ϕ ≥ 2 and 1 ∈ D(ϕ).
Case 1: ϕK is isotropic.
Let x ∈ D(ϕK ). Suppose that K = F (x). Let p ∈ F [t] denote the (monic) minimal
polynomial
of®x so K = F (p). It follows from the Quadratic Value Theorem 18.3 that
p ∈ D(ϕF (t) ) and deg p is even. In particular, NK/F (x) = p(0) and by the Substitution
Principle 17.7, ®
NK/F (x) = p(0) ∈ D(ϕ) .
®
If F (x) ( K let m = [K : F (x)]. If m is even then NK/F (x) ∈ F ×2 ⊂ D(ϕ) . If m is
odd then ϕF (x) is isotropic by Springer’s Theorem 18.5. Applying the above argument to
the field extension F (x)/F yields
®
NK/F (x) = NF (x)/F (x)m ∈ D(ϕ)
as needed.
Case 2: ϕK is anisotropic.
Let x ∈ D(ϕK ). Choose vectors v, v0 ∈ VK such that ϕK (v) = x and ϕK (v0 ) = 1. Let V 0 ⊂
VK be a 2-dimensional subspace (over K) containing v and v0 . The restriction ϕ0 of ϕK
to V 0 is a binary anisotropic quadratic form over K representing x and 1. It follows from
Proposition 12.1 that the even Clifford algebra L = C0 (ϕ0 ) is a quadratic field extension
19. FORMS OVER A DISCRETE VALUATION RING 75
of K and x = NL/K (y) for some y ∈ L× . Moreover, since C0 (ϕ0L ) = C0 (ϕ0 ) ⊗K L = L ⊗K L

is not a field, by the same proposition, ϕ0 and therefore ϕ is isotropic over L. Applying
Case 1 to the field extension L/F yields
¡ ¢ ®
NK/F (x) = NK/F NL/K (y) = NL/F (y) ∈ D(ϕ) . ¤
Theorem 18.11. (Bilinear Value Norm Principle) Let b¡be a symmetric
¢ bilinear
® form
over F and let K/F be a finite field extension. Then NK/F D(bK ) ⊂ D(b) .
Proof. As D(bE ) = D(ϕbE ) for any field extension E/F , this follows from the qua-
dratic version of the theorem. ¤
19. Forms over a discrete valuation ring
We wish to look at similarity factors of bilinear and quadratic forms. To do so we

need a few facts about such forms over a discrete valuation ring (DVR) which we now
establish. These results are based on the work of Springer in [131] in the case of fields of
characteristic different from two.
Throughout this section, R will be a DVR with quotient field K, residue field K̄, and
prime element π. If V is a free R-module of finite rank then the definition of a (symmetric)
bilinear form and quadratic form on V is analogous to the field case. In particular, we
can associate to every quadratic form its polar form bϕ : (v, w) 7→ ϕ(v + w) − ϕ(v) −
ϕ(w). Orthogonal complements are defined in the usual way. Orthogonal sums of bilinear
(respectively, quadratic) forms are defined as in the field case. We use analogous notation
as in the field case when clear. If F → R is a ring homomorphism and ϕ is a quadratic
form over F , we let ϕR = R ⊗F ϕ.
A bilinear form b on V is non-degenerate if l : V → HomR (V, R) defined by v 7→ lv :
w → b(v, w) is an isomorphism. As in the field case, we have the crucial
Proposition 19.1. Let R be a DVR. Let V be a free R-module of finite rank and
W a submodule of V . If ϕ is a quadratic form on V with bϕ |W non-degenerate then
ϕ = ϕ|W ⊥ ϕ|W ⊥ .
Proof. As bϕ |W is non-degenerate, W ∩ W ⊥ = {0} and if v ∈ V there exists w0 ∈ W
such that the linear map W → F by w 7→ bϕ (v, w) is given by bϕ (v, w) = bϕ (w0 , w) for
all w ∈ W . Consequently, v = w + (v − w0 ) ∈ W ⊕ W ⊥ and the result follows. ¤
Hyperbolic quadratic forms and planes are also defined in an analogous way. We let
H denote the quadratic hyperbolic plane.
If R is a DVR and V a vector space over the quotient field K of R. A vector v ∈ V is
called primitive if it is not divisible by a prime element π, i.e., the image v̄ of v in K̄ ⊗R V
is not zero.
Arguing as in Proposition 7.14, we have
Lemma 19.2. Let R be a DVR. Let ϕ be a quadratic form on V whose polar form
is non-degenerate. Suppose that V contains an isotropic vector v. Then there exists a
submodule W of V containing v such that ϕ|W ' H.
Proof. Dividing v by π n for an appropriate choice of n, we may assume that v is

primitive. It follows easily that V /Rv is torsion-free hence free. In particular, V → V /Rv
splits, therefore, Rv is a direct summand of V . Let f : V → R be an R-linear map
satisfying f (v) = 1. As l : V → HomR (V, R) is an isomorphism, there exists an element
w ∈ V such that f = lw hence bϕ (v, w) = 1. Let W = Rv ⊕ Rw. Then v, w − ϕ(w)v is a
hyperbolic pair. ¤
By induction, we conclude:
Corollary 19.3. Suppose that R is a DVR. Let ϕ be a quadratic form on V over R
whose polar form is non-degenerate. Then ϕ = ϕ|V1 ⊥ ϕ|V2 with V1 , V2 submodules of V
satisfying ϕ|V1 is anisotropic and ϕ|V2 ' mH for some m ≥ 0.
Associated to a quadratic form ϕ on V over R are two forms: ϕK on K ⊗R V over K

and ϕ̄ = ϕK̄ on K̄ ⊗R V over K̄.
Lemma 19.4. Suppose that R be a complete DVR. Let ϕ be an anisotropic qua-
dratic form over R whose associated bilinear form bϕ is non-degenerate. Then ϕ̄ is also
anisotropic.
Proof. Let {v1 , . . . , vn } be a basis for Vϕ and t1 , . . . , tn the respective coordinates.
∂ϕ
If w ∈ Vϕ then (w) = bϕ (vi , w). In particular, if w̄ 6= 0̄ there exists an i such that
∂ti
b¯ϕ (v̄i , w̄) 6= 0. It follows by Hensel’s lemma that ϕ would be isotropic if ϕ̄ is. ¤
Lemma 19.5. Let ϕ and ψ be two quadratic forms over a DVR R such that ϕ̄ and ψ̄
are anisotropic over K̄. Then ϕK ⊥ πψK is anisotropic over K.
Proof. Suppose that ϕ(u) + πψ(v) = 0 for some u ∈ Vϕ and v ∈ Vψ with at least
one of u and v primitive. Reducing modulo π, we have ϕ̄(ū) = 0. Since ϕ̄ is anisotropic,
u = πw for some w. Therefore πϕ(w) + ψ(v) = 0 and reducing modulo π we get ψ̄(v̄) = 0.
Since ψ̄ is also anisotropic, v is divisible by π, a contradiction. ¤
Corollary 19.6. Let ϕ and ψ be an anisotropic forms over F . Then ϕF (t) ⊥ tψF (t)
is anisotropic.
Proof. In the lemma, let R = F [t](t) , a DVR, π = t a prime. As ϕR = ϕ and ψR = ψ,
the result follows from the lemma. ¤
Proposition 19.7. Let ϕ be a quadratic form over a complete DVR R such that
the associated bilinear form bϕ is non-degenerate. Suppose that ϕK ' πϕK . Then ϕ̄ is
hyperbolic.
Proof. Write ϕ = ψ ⊥ nH with ψ anisotropic. By Lemma 19.4, we have ψ̄ is
anisotropic. The form
ϕK ⊥ (−πϕK ) ' ψK ⊥ (−πψK ) ⊥ 2nH
is hyperbolic and ψK ⊥ (−πψK ) is anisotropic over K by Lemma 19.5. We must have
ψ = 0 by uniqueness of the Witt decomposition over K, hence ϕ = nH is hyperbolic. It
follows that ϕ̄ is hyperbolic. ¤
19. FORMS OVER A DISCRETE VALUATION RING 77
Proposition 19.8. Let ϕ be a non-degenerate quadratic form over F of even dimen-

sion. Let f ∈ F [T ] and p ∈ F [T ] an irreducible polynomial factor of f of odd multiplicity.
If ϕF (T ) ' f ϕF (T ) then ϕF (p) is hyperbolic.
Proof. Let R denote the completion of the DVR F [T ](p) and let K be its quotient
field. The residue field of R coincides with F (p). Modifying f by a square, we may assume
that f = up for some u ∈ R× . As ϕF (T ) ' f ϕF (T ) , we have ϕF (T ) ' upϕF (T ) . Applying
Proposition 19.7 to the form ϕR and π = up yields (ϕR ) = ϕF (p) is hyperbolic. ¤
We shall also need the following:

Proposition 19.9. Let R be a DVR with quotient field K. Let ϕ and ψ be two
quadratic forms on V and W over R respectively such that their respective residues forms
ϕ̄ and ψ̄ are anisotropic. If ϕK ' ψK then ϕ ' ψ (over R).
Proof. Let f : VK → WK be an isometry between ϕK and ψK . It suffices to prove
that f (V ) ⊂ W and f −1 (W ) ⊂ V . Suppose that there exists a v ∈ V such that f (v)
is not in W . Then f (v) = w/π k for some primitive w ∈ W and k > 0. Since f is an
isometry, we have ψ(w) = π 2k ϕ(v), i.e., ψ(w) is divisible by π, hence w̄ is an isotropic
vector of ψ̄, a contradiction. Analogously, f −1 (W ) ⊂ V . ¤
19.A. Residue homomorphisms. If R is a DVR then for each x ∈ K × we can
write x = uπ n for some u ∈ R× and n ∈ Z.
Lemma 19.10. Let R be a DVR with quotient field K and residue field K̄. Let π be a
prime element in R. There exist group homomorphisms
∂ : W (K) → W (K̄) and ∂π : W (K) → W (K̄)
satisfying
( (
¡ ¢ hūi n is even ¡ ¢ hūi n is odd
∂ huπ n i = and ∂π huπ n i =
0 n is odd 0 n is even
for u ∈ R× and n ∈ Z.
Proof. It suffices to prove the existence of ∂ as we can take ∂π = ∂ ◦ λπ where λπ is
the group homomorphism λπ : W (K) → W (K) given by b → πb.
By Theorem 4.8, it suffices to check the generating relations of the Witt ring are
respected. As h1i + h−1i = 0 in W (K̄), it suffices to show if a, b ∈ R with a + b 6= 0 then
¡ ¢ ¡ ¢ ¡ ¢ ¡ ¢
(19.11) ∂ hai + ∂ hbi = ∂ ha + bi + ∂ hab(a + b)i
in W (K̄).
Let
a = a0 π n , b = b0 π m a + b = π l c0 with a0 , b0 , c0 ∈ R×
and m, n, l ∈ Z satisfying min{m, n} ≤ l. We may assume that n ≤ m.
Suppose that n < m. Then
b0 b0
a + b = π n a0 (1 + π m−n ) and ab(a + b) = π 2n+m b0 a20 (1 + π m−n ).
a0 a0
¡ ¢ ¡ ¢ ¡ ¢ ¡ ¢
In particular, ∂ hai = ∂ ha + bi and ∂ hbi = ∂ hab(a + b)i as needed.
Suppose that n = m.
If n = l then a0 + b0 ∈ R× and the result follows by the Witt relation in W (K̄).
So suppose
¡ that¢ n < l. Then ¡ ā0 = −b̄¢0 so the left hand side of (19.11) is zero. If l is
odd then ∂ ha + bi = 0 = ∂ hab(a + b)i as needed. So we may assume that l is even.
Then ha + bi ' hc0 i and hab(a + b)i ' ha0 b0 c0 i over K. Hence the right hand side of
(19.11) is hc̄0 i + hā0 b̄0 c̄0 i = hc̄0 i + h−c̄0 i = 0 in W (K̄) also. ¤
The map ∂ : W (K) → W (K̄) in the lemma does not dependent on the choice on the
prime element π. It is called the first residue homomorphism with respect to R. The map
∂π : W (K) → W (K̄) does depend on π. It is called the second residue homomorphism
with respect to R and π.
Remark 19.12. Let R be a DVR with quotient field K and residue field K̄. Let π be
a prime element in R. If b is a non-degenerate diagonalizable bilinear form over K, we
can write b as
b ' hu1 , . . . , un i ⊥ πhv1 , . . . , vm i
×
for some ui , vj ∈ R . Then ∂(b) = hū1 , . . . , ūn i in W (K̄) and ∂π (b) = hv̄1 , . . . , v̄m i in
W (K̄).
Example 19.13. Let R be a DVR with quotient field K and residue field K̄. Let π
be a prime element in R. Let b = hha1 , . . . , an ii, an anisotropic n-fold Pfister form over
K. Then we may assume that ai = π ji ui with ji = 0 or 1 and ui ∈ R× for all i. By
Corollary 6.13, we may assume that ai ∈ R× for all i > 1. As b = −a1 hha2 , . . . , an ii ⊥
hha2 , . . . , an ii, if a1 ∈ R× then ∂(b) = hhā1 , . . . , ān ii and ∂π (b) = 0, and if a1 = πu1 then
∂(b) = hhā2 , . . . , ān ii and ∂π (b) = −ū1 hhā2 , . . . , ān ii.
As n-fold Pfister forms generate I n (F ), we have, by the example, the following:
Lemma 19.14. Let R be a DVR with quotient field K and residue field K̄. Let π be a
prime element in R. Then for every n ≥ 1:
¡ ¢
(1) ∂ ¡I n (K) ¢⊂ I n−1 (K̄).
(2) ∂π I n (K) ⊂ I n−1 (K̄).
Exercise 19.15. Suppose that R is a complete DVR with quotient field K and residue
field K̄. If char K̄ 6= 2 then the residue homomorphisms induce split exact sequences of
groups:
0 → W (K̄) → W (K) → W (K̄) → 0
and
0 → I n (K̄) → I n (K) → I n−1 (K̄) → 0.
20. Similarities of forms
Let ϕ be an anisotropic quadratic form over F . Let p ∈ F [T ] := F [t1 , . . . , tn ] be

irreducible and F (p) the quotient field of F [T ]/(p). In this section, we determine what it
means for ϕF (p) to be hyperbolic. We also establish the analogous result for anisotropic
bilinear forms over F . We saw that a form to becoming isotropic over F (p) was related to
20. SIMILARITIES OF FORMS 79
the values it represented over the polynomial ring F [T ]. We shall see that hyperbolicity
is related to the similarity factors of the form over F [T ]. We shall also deduce norm
principles for similarity factors of a form over F first established by Scharlau in [123]
and Knebusch in [82]. To establish these results, we introduce the transfer of forms from
a finite extension of F to F , an idea introduced by Scharlau (cf. [123]) for quadratic
forms over fields of characteristic different from two and Baeza (cf. [17]) in arbitrary
characteristic.
20.A. Transfer of bilinear and quadratic forms. Let K/F be a finite field ex-
tension and s : K → F an F -linear functional. If b is a symmetric bilinear form on V
over K define the transfer s∗ (b) of b induced by s to be the symmetric bilinear form on
V over F given by ¡ ¢
s∗ (b)(v, w) = s b(v, w) for all v, w ∈ V.
If ϕ is a quadratic form on V over K define the transfer
¡ s¢∗ (ϕ) of ϕ induced by s to be the
quadratic form on V over F given by s∗ (ϕ)(v) = s ϕ(v) for all v ∈ V with polar form
s∗ (bϕ ).
Note that dim s∗ (b) = [K : F ] dim b.
Lemma 20.1. Let K/F be a finite field extension and s : K → F an F -linear func-
tional. The transfer s∗ factors through orthogonal sums and preserves isometries.
¡ ¢
Proof. Let v, w ∈ Vb . If b(v, w) = 0 then s∗ (b)(v, w) = s b(v, w) = 0. Thus
s∗ (b ⊥ c) = s∗ (b) ⊥ s∗ (c). If σ : b → b0 is an isometry then
¡ ¢ ¡ ¢ ¡ ¢
s∗ (b0 ) σ(v), σ(w) = s b0 (σ(v), σ(w)) = s b(v, w) = s∗ (b)(v, w),
so σ : s∗ (b) → s∗ (b0 ) is also an isometry. ¤
Proposition 20.2. (Frobenius Reciprocity) Suppose that K/F is a finite extension
of fields and s : K → F an F -linear functional. Let b and c be symmetric bilinear forms
over F and K respectively and ϕ and ψ quadratic forms over F and K respectively. Then
there exist canonical isometries:
(20.3a) s∗ (bK ⊗K c) ' b ⊗F s∗ (c).

(20.3b) s∗ (bK ⊗K ψ) ' b ⊗F s∗ (ψ).
(20.3c) s∗ (c ⊗K ϕK ) ' s∗ (c) ⊗F ϕ.
In particular, ¡ ¢
s∗ (bK ) ' b ⊗F s∗ h1ib .
Proof. (a): The canonical F -linear map VbK ⊗K Vc → Vb ⊗F Vc given by (a⊗v)⊗w 7→
v ⊗ aw is an isometry. Indeed
¡ ¢ ¡ ¢
s(bK ⊗ c) (a ⊗ v) ⊗ w, (a0 ⊗ v 0 ) ⊗ w0 = s aa0 b(v, v 0 )c(w, w0 )
¡ ¢
= b(v, v 0 )s c(aw, a0 w0 ) = (b ⊗ sc)(v ⊗ aw, v 0 ⊗ a0 w0 ).
The last statement follows from the first by setting c = h1i.
(b) and (c) are proved in a similar fashion. ¤
Lemma 20.4. Let K/F be a finite field extension and s : K → F a nonzero F -linear
functional.
(1) If b is a non-degenerate symmetric bilinear form on V over K then s∗ (b) is
non-degenerate on V over F .
(2) If ϕ is an even dimensional non-degenerate quadratic form on V over K then
s∗ (ϕ) is non-degenerate on V over F .
Proof. Suppose that 0 6= v ∈ V . As b is non-degenerate, there exists a w ∈ V such
that 1 = b(v, w). As s is not zero, there exists a c ∈ K such that 0 6= s(c) = s∗ (b)(v, cw).
This shows (1). Statement (2) follows from (1) and Remark 7.22(1). ¤
Corollary 20.5. Let K/F be a finite extension of fields and s : K → F a nonzero
F -linear functional.
(1) If c is a bilinear hyperbolic form over K then s∗ (c) is a hyperbolic form over F .
(2) If ϕ is a quadratic hyperbolic form over K then s∗ (ϕ) is a hyperbolic form over
F.
Proof. (1): As s∗ respects orthogonality, we may assume that c = H1 . By Frobenius
Reciprocity, ¡ ¢ ¡ ¢
s∗ (H1 ) ' s∗ (H1 )K ' (H1 )F ⊗ s∗ h1i .
¡ ¢
As s∗ h1i is non-degenerate by Lemma 20.4, we have s∗ (H1 ) is hyperbolic by Lemma 2.1.
(2): This follows in the same way as (1) using Lemma 8.16. ¤
Definition 20.6. Let K/F be a finite field extension and s : K → F a nonzero
F -linear functional. By Lemmas 20.4 and 20.5, the functional s induces group homomor-
phisms
c (K) → W
s∗ : W c (F ), s∗ : W (K) → W (F ), and s∗ : Iq (K) → Iq (F )
called transfer maps. Let b and c be non-degenerate symmetric bilinear form over F and
K respectively and ϕ and ψ non-degenerate quadratic forms over F and K respectively.
By Frobenius Reciprocity, we have
¡ ¢
s∗ rK/F (b) · c = b · s∗ (c)
c (F ) and W (F ), i.e., s∗ : W
in W c (K) → W c (F ) is a W
c (F )-module homomorphism and
s∗ : W (K) → W (F ) is a W (F )-module homomorphism where we view W (K) as a
W (F )-module via rK/F . Furthermore,
¡ ¢ ¡ ¢
s∗ rK/F (b) · ψ = b · s∗ (ψ) and s∗ c · rK/F (ϕ) = s∗ (c) · ϕ
¡ ¢
in Iq (F ). Note that s∗ I(K) ⊂ I(F ).
Corollary 20.7. Let K/F be a finite field extension and s : K → F a nonzero
F -linear functional. Then the compositions
c (F ) → W
s∗ rK/F : W c (F ), s∗ rK/F : W (F ) → W (F ), and s∗ rK/F : Iq (F ) → Iq (F )
¡ ¢ ¡ ¢
are given by multiplication by s∗ h1i
¡ b , ¢i.e., b →
7 b · s∗ h1i b for a non-degenerate sym-
metric bilinear form b and ϕ 7→ s∗ h1ib · ϕ for a non-degenerate quadratic form ϕ.
Corollary 20.8. Let K/F be a field extension and s : K → F a nonzero F -linear

c (F ) (respectively, W (F )) and is independent of
functional. Then im(s∗ ) is an ideal in W
s.
Proof. By Frobenius Reciprocity, im(s∗ ) is an ideal. Suppose that s1 : K → F is
another nonzero¡ F -linear
¢ functional. Let K → HomF (K, F ) be the F -isomorphism given
by a 7→ (x 7→ s ax) . Hence there exists a unique a ∈ K × such that s1 (x) = s(ax) for all
x ∈ K. Consequently, (s1 )∗ (b) = s∗ (ab) for all non-degenerate symmetric bilinear forms
b over K. ¤
Let K = F (x)/F be an extension of degree n and a = NK/F (x) ∈ F × the norm of x.

Let
s : K → F be the F -linear functional defined by
(20.9)
s(1) = 1 and s(xi ) = 0 for all i = 1, . . . , n − 1.
Then s(xn ) = (−1)n+1 a.
Lemma 20.10. The transfer induced by the F -linear functional s in (20.9) satisfies
½
¡ ¢ h1ib if n is odd
s∗ h1ib =
h1, −aib if n is even.
¡ ¢
Proof. Let b = s∗ h1i . Let V ⊂ K be the F -subspace spanned by xi with i =
1, . . . , n, a non-degenerate subspace. Then V ⊥ = F , consequently K = F ⊕ V .
First suppose that n = 2m+1 is odd. The subspace of W spanned by xi , i = 1, . . . , m,
is a lagrangian of b|V , hence b|V is metabolic and b = b|V ⊥ = h1i in W (F ).
Next suppose that n = 2m is even. We have
½
i j 0 if i + j < n
b(x , x ) =
−a if i + j = n.
It follows that det b = (−1)m aF ×2 and the subspace W 0 ⊂ W spanned by all xi with
i 6= m and 1 ≤ i ≤ n is non-degenerate. In particular, K = W 0 ⊕ (W 0 )⊥ by Proposition
1.7. By dimension count dim(W 0 )⊥ = 2. As the subspace of W 0 spanned by xi , i =
1, . . . , m − 1, is a lagrangian of b|W 0 , we have b|W 0 is metabolic. Computing determinants,
yields b|(W 0 )⊥ ' h1, −ai, hence in W (F ), we have b = b|(W 0 )⊥ = h1, −ai. ¤
Corollary 20.11. Suppose ¡ that K = F¢ (x) is a finite extension of even degree over
F . Then ker(rK/F ) ⊂ annW (F ) hhNK/F (x)ii .
Proof. Let s be the F -linear functional in (20.9). By Corollary 20.7 and Lemma
20.10, we have
¡ ¢ ¡ ¢ ¡ ¢
ker rK/F : W (F ) → W (K) ⊂ annW (F ) s∗ (h1i) = annW (F ) hhNK/F (x)ii . ¤
Corollary 20.12. Let K/F be a finite field extension of odd degree. Then the map
rK/F : W (F ) → W (K) is injective.
Proof. If K = F (x) and s is as in (20.9) then by Corollary 20.7 and Lemma 20.10,
we have
¡ ¢ ¡ ¢ ¡ ¢
ker rK/F : W (F ) → W (K) ⊂ annW (F ) s∗ (h1i) = annW (F ) h1i = 0.
The general case follows by induction of the odd integer [K : F ]. ¤
Note that this corollary provides a more elementary proof of Corollary 18.6.
Lemma 20.13. The transfer induced by the F -linear functional s in (20.9) satisfies
½
¡ ¢ haib if n is odd
s∗ hxib =
0 if n is even.
¡ ¢
Proof. Let b = s∗ hxi . First suppose that n = 2m + 1 is odd. Then
½
i j 0 if i + j < n − 1
b(x , x ) =
a if i + j = n − 1.
It follows that det b = (−1)m aF ×2 and the subspace W ⊂ K spanned by all xi with i 6= m
and 1 ≤ i ≤ n is non-degenerate. In particular, K = W ⊕ W ⊥ by Proposition 1.7 and W ⊥
is 1-dimensional by dimension count. Computing determinants, we see that b|W ⊥ ' hai.
As the subspace of W spanned by xi , i = 0, . . . , m − 1, is a lagrangian of b|W , the form
b|W is metabolic. Consequently, b = b|W ⊥ = hai in W (F ).
Next suppose that n = 2m is even. The subspace of K spanned by xi , i = 0, . . . , m−1,
is a lagrangian of b so b is metabolic and b = 0 in W (F ). ¤
Corollary¡ 20.14.¢ Let s∗ be the transfer induced by the F -linear functional s in
(20.9). Then s∗ hhxii = hhaii in W (F ).
20.B. Similarity theorems. As a consequence, we get the norm principle first es-
tablished by Scharlau in [123].
Theorem 20.15. (Similarity Norm Principle) Let K/F be a finite field extension and
ϕ a non-degenerate even dimensional quadratic form over F . Then
¡ ¢
NK/F G(ϕK ) ⊂ G(ϕ).
Proof. Let x ∈ G(ϕK ). Suppose first that K = F (x). Let s be as in (20.9). As
hhxii · ϕK = 0 in Iq (K), applying the transfer s∗ : Iq (K) → Iq (F ) yields
¡ ¢ ¡ ¢
0 = s∗ hhxii · ϕK = s∗ hhxii · ϕ = hhNK/F (x)ii · ϕ
in Iq (F ) by Frobenius Reciprocity 20.2 and Corollary 20.14. Hence NK/F (x) ∈ G(ϕ) by
Remark 8.17.
In the general case, set k = [K : F (x)]. If k is even we have
NK/F (x) = NF (x)/F (x)k ∈ G(ϕ)
¡ ¢
since F ×2 ⊂ G(ϕ). If k is odd, the homomorphism Iq F (x) → Iq (K) is injective by
Remark 18.6, hence hhxii · ϕF (x) = 0. By the first part of the proof, NF (x)/F (x) ∈ G(ϕ).
Therefore, NK/F (x) ∈ NF (x)/F (x)F ×2 ⊂ G(ϕ). ¤
We turn to similarities of forms over polynomial rings. As with values, Knebusch

proved analogous results for similarities in [82].
Lemma 20.16. Let ϕ be a non-degenerate quadratic form of even dimension and p ∈
F [t] a monic irreducible polynomial (in one variable). If ϕF (p) is hyperbolic then p ∈
G(ϕF (t) ).
Proof. Let x be the image of t in K = F (p) = F [t]/(p). We have p is the norm of
t − x in the extension K(t)/F (t). Since ϕK(t) is hyperbolic, t − x ∈ G(ϕK(t) ). Applying
the Norm Principle 20.15 to the form ϕF (t) and the field extension K(t)/F (t) yields
p ∈ G(ϕF (t) ). ¤
Theorem 20.17. (Quadratic Similarity Theorem) Let ϕ be a non-degenerate quadratic
form of even dimension and f ∈ F [T ] = F [t1 , . . . , tn ] a nonzero polynomial. Then the
following conditions are equivalent:
(1) f ∗ f ∈ G(ϕF (T ) ).
(2) There exists an a ∈ F × such that af ∈ G(ϕF (T ) ).
(3) For any irreducible divisor p of f to an odd power, the form ϕF (p) is hyperbolic.
(2) ⇒ (3) follows from Proposition 19.8.
(3) ⇒ (1): We proceed by induction on the number n of variables. We may assume that f
is irreducible and degt1 f > 0. In particular, f is an irreducible polynomial in t1 over the
field E = F (T 0 ) = F (t2 , . . . , tn ). Let g ∈ F [T 0 ] be the leading term of f . In particular,
g ∗ = f ∗ . As the polynomial f 0 = f g −1 in E[t1 ] is monic irreducible and E(f 0 ) = F (f ),
the form ϕE(f 0 ) is hyperbolic. Applying Lemma 20.16 to ϕE and the polynomial f 0 , we
have f g = f 0 · g 2 ∈ G(ϕF (T ) ).
Let p ∈ F [T 0 ] be an irreducible divisor of g to an odd power. Since p does not divide f ,
by the first part of the proof ¡applied ¢ to the ¡ polynomial¢ f g, the form ϕF (p)(t1 ) is hyperbolic.
Since the homomorphism Iq F (p) → Iq F (p)(t1 ) is injective by Remark 8.18, we have
ϕF (p) is hyperbolic. Applying the induction hypothesis to g yields g ∗ g ∈ G(ϕF (T 0 ) ).
Therefore, f ∗ f = g ∗ f = g ∗ g · f g · g −2 · ∈ G(ϕF (T ) ). ¤
Theorem 20.18. (Bilinear Similarity Norm Principle) Let K/F be a finite field ex-
tension and b an anisotropic symmetric bilinear form over F of positive dimension. Then
¡ ¢
NK/F G((bK )an ) ⊂ G(b).
¡ ¢
Proof. Let x ∈ G (bK )an . Suppose first that K = F (x). Let s be as in (20.9). Let
bK = (bK )an ⊥ c with c a metabolic form over K. Then xc is metabolic so
¡ ¢
bK = (bK )an = x(bK )an = x (bK )an + c = xbK
in W (K). Consequently, hhxii·bK = 0 in I(K). Applying the transfer s∗ : W (K) → W (F )
yields ¡ ¢ ¡ ¢
0 = s∗ hhxii · bK = s∗ hhxii · b = hhNK/F (x)ii · b
by Frobenius Reciprocity 20.2 and Corollary 20.14. Hence NK/F (x)b = b in W (F ) with
both sides anisotropic. It follows from Proposition 2.4 that NK/F (x) ∈ G(b).
In the general case, set k = [K : F (x)]. If k is even we have

NK/F (x) = NF (x)/F (x)k ∈ G(b)
¡ ¢
since F ×2 ⊂ G(b). If k is odd, the homomorphism¡ W F¢(x) → W (K) is ¡ injective¢
by Corollary 18.6, hence hhxii · (bF (x) )an = 0 in W F (x) . Thus x ∈ G (bF (x) )an
by Proposition 2.4. By the first part of the proof, NF (x)/F (x) ∈ G(b). Consequently,
NK/F (x) ∈ NF (x)/F (x)F ×2 ⊂ G(b). ¤
Lemma 20.19. Let b be a non-degenerate anisotropic symmetric bilinear form and
p ∈ F [t] a monic irreducible polynomial (in one variable). If bF (p) is metabolic then
p ∈ G(bF (t) ).
Proof. Let x be the image of t in K = F (p) = F [t]/(p). We have p is the norm
of t − x in¡ the extension
¢ K(t)/F (t). Since bK(t) is metabolic, (bK(t) )an = 0. Thus
x − t ∈ G (bK(t) )an . Applying the Norm Principle 20.18 to the anisotropic form bF (t)
and the field extension K(t)/F (t) yields p ∈ G(bF (t) ). ¤
Theorem 20.20. (Bilinear Similarity Theorem) Let b be an anisotropic bilinear form
of even dimension and f ∈ F [T ] = F [t1 , . . . , tn ] a nonzero polynomial. Then the following
conditions are equivalent:
(1) f ∗ f ∈ G(bF (T ) ).
(2) There exists an a ∈ F × such that af ∈ G(bF (T ) ).
(3) For any irreducible divisor p of f to an odd power, the form bF (p) is metabolic.
Proof. Let ϕ = ϕb be of dimension m.
(1) ⇒ (2) is trivial.
(2) ⇒ (3): Let p be an irreducible factor of f to an odd degree. As F (T ) is the quotient
field of¡ the localization
¢ ¡ F¢[T ](p) and F [T ](p) is a DVR, we have a group homomorphism
∂ : W F (T ) → W F (p) given by Lemma 19.10. Since p is a divisor to an odd power
of f ,
bF (p) = ∂(bF (T ) ) = ∂(af bF (T ) ) = 0
¡ ¢
in W F (p) . Thus bF (p) is metabolic.
(3) ⇒ (1): The proof is analogous to the proof of (3) ⇒ (1) in the Quadratic Similarity
Theorem 20.17 with Lemma 20.19 replacing Lemma 20.16 and hyperbolicity replaced by
metabolicity. ¤
Corollary 20.21. Let ϕ be an quadratic form (respectively, b an anisotropic bilinear
form) on V over F ¡and f ¢∈ F [T ] with T = (t1 , . . . , tn ). Suppose that f ∈ G(ϕF (T ) )
(respectively, f ∈ G bF (T ) ) . If f (a) is defined and nonzero with a ∈ F n then f (a) ∈
G(ϕ).
Proof. We may assume that ϕ is anisotropic as G(ϕ) = G(ϕan ). (Cf. Remark 8.9.)
By induction, we may assume that f is a polynomial in one variable t. Let R = F [t](t−a) , a
DVR. As f (a) 6= 0, we have f ∈ R× . Over F (t), we have ϕF (t) ' f ϕF (t) hence ϕR ' f ϕR
by Proposition 19.9. Since F is the residue class field of R, upon taking the residue forms,
we see that ϕ = f (a)ϕ as needed.

21. AN EXACT SEQUENCE FOR W F (t) 85
As in the quadratic case, we reduce to f being a polynomial in one variable. We then

have bF (t) ' f bF (t) Taking ∂ of this equation relative to the DVR R = F [t](t−a) yields
b̄ = f¯b̄ = f (a)b̄ in W (F ) as f ∈ R× . The result follows by Proposition 2.4. ¤
Corollary 20.22. Let ϕ be an quadratic form (respectively, b an anisotropic bilinear
form) on V over F and g ∈ F [T ]. Suppose that g ∈ G(ϕF (T ) ) (respectively, g ∈ G(bF (T ) ).
Then g ∗ ∈ G(ϕ) (respectively, g ∗ ∈ G(b)).
Proof. We may assume that ϕ is anisotropic as G(ϕ) = G(ϕan ). (Cf. Remark 8.9.)
By induction on the number of variables, we may assume that g ∈ F [t]. By Lemma 18.1
and Lemma 9.2, we must have deg g = 2r is even. Let h(t) = t2r g(1/t) ∈ G(ϕF (t) ). Then
g ∗ = h(0) ∈ G(ϕ) by Corollary 20.21. An analogous proof shows the result for symmetric
bilinear forms (using also Lemma 9.3 to see that deg g is even). ¤
¡ ¢
21. An exact sequence for W F (t)
Let A1F be the one dimensional affine line over F . Let x ∈ A1F be a closed point and
F (x) be the residue field of x. Then there exists a unique monic irreducible polynomial
fx ∈ F [t] of degree d = deg x such that F (x) = F [t]/(fx ). By Lemma 19.10, we have
the first and second residue homomorphisms with respect to the DVR OA1F ,x and prime
element fx :
¡ ¢ ∂ ¡ ¢ ¡ ¢ ∂fx ¡ ¢
W F (t) − → W F (x) and W F (t) −−→ W F (x) .
¡ ¢
Denote ∂fx by ∂x . If g ∈ F [t] then ∂x hgi = 0 unless fx | g in F [t]. It follows if b is a
non-degenerate bilinear form over F (t) that ∂x (b) = 0 for almost all x ∈ A1F .
We have
Theorem 21.1. The sequence
rF (t)/F ¡ ¢ @ a ¡ ¢
0 → W (F ) −−−−→ W F (t) −→ W F (x) → 0
x∈A1F
is split exact where ∂ = (∂x ).

Proof. As anisotropic bilinear forms remain anisotropic under a purely transcenden-
tal extension, rF (t)/F is monic. It is split by the first residue homomorphism with respect
to any rational point in A1F .
¡ ¢
Let F [t]d := {g | g ∈ F [t], deg g ≤ d} and Ld ⊂ W F¡(t) the ¢ subring generated
by hgi with g ∈ F [t]d . Then L0 ⊂ L1 ⊂ L2 ⊂ · · · and W F (t) = ∪d Ld . Note that
im(rF (t)/F ) = L0 . Let Sd be the multiplicative monoid in F [t] generated by F [t]d \ {0}.
As a group Ld is generated by one-dimensional forms of the type
(21.2) hf1 · · · fm gi
with distinct monic irreducible polynomials f1 , . . . , fm ∈ F [t] of degree d and g ∈ Sd−1 .
Claim: The additive group Ld /Ld−1 is generated by hf gi + Ld−1 with f ∈ F [t] monic
irreducible of degree d and g ∈ Sd−1 . Moreover, if h ∈ F [t]d−1 satisfies g ≡ h mod (f )
then hf gi ' hf hi mod Ld−1 .
We first must show that a generator of the form in (21.2) is a sum of the desired forms
mod Ld−1 . By induction on m, we need only do the case m = 2. Let f1 , f2 be distinct
irreducible monic polynomials of degree d and g ∈ Sd−1 . Let h = f1 − f2 so deg h < d.
We have
hf1 i = hhi + hf2 i − hf1 f2 hi
¡ ¢
in W F (t) by the Witt relation (4.2). Multiplying this equation by hf2 gi and deleting
squares, yields
hf1 f2 gi = hf2 ghi + hgi − hf1 ghi ≡ hf2 ghi − hf1 ghi mod Ld−1
as needed.
Now suppose that g = g1 g2 with g1 , g2 ∈ F [t]d−1 . As f 6 | g by the Division Algorithm,
there exist polynomials q, h ∈ F [t] with h 6= 0 and deg h < d satisfying g = f q + h. It
follows that deg q < d. By the Witt relation (4.2), we have
hgi = hf qi + hhi − hf qhgi
in Ld hence multiplying by hf i, we have
hf gi = hqi + hf hi − hqhgi ≡ hf hi mod Ld−1 .
The Claim now follows by induction on the number of factors for a general g ∈ Sd−1 .
Let x ∈ A1F be of degree d and f = fx . Define
¡ ¢
αx : W F (x) → Ld /Ld−1 by hg + (f )i 7→ hgi + Ld−1 for g ∈ F [t]d−1 .
We show this map is well-defined. If h ∈ F [t]d−1 satisfies gh2 ≡ l mod (f ), with l ∈
F [t]d−1 then hf gi = hf gh2 i ≡ hf li mod Ld−1 by the Claim, so the map is well-defined on
1-dimensional forms. If g1 , g2 ∈ F [t]d−1 satisfy g1 + g2 6= 0 and h ≡ (g1 + g2 )g1 g2 mod (f )
then
hf g1 i + hf g2 i = hf (g1 + g2 )i + hf g1 g2 (g1 + g2 )i ≡ hf (g1 + g2 )i + hf hi mod Ld−1
¡ ¢
by the Claim. As hf i+h−f i = 0 in W F (t) , it follows that αx is well-defined by Theorem
4.8.
Let x0 ∈ A1F with deg x0 = d. Then the composition
¡ ¢ αx ∂x0 ¡ ¢
W F (x) −→ Ld /Ld−1 −→ W F (x0 )
is the identity if x = x0 otherwise it is the zero map. It follows that the map
a ¡ ¢ (αx )
W F (x) −−→ Ld /Ld−1
deg x = d
is split by (∂x )deg x=d . As (αx ) is surjective by the Claim, it is an isomorphism with inverse
(∂x )deg x=d . By induction on d, we check that
a ¡ ¢
(∂x )deg x≤d : Ld /L0 −→ W F (x)
deg x≤d
is an isomorphism. As L0 = W (F ), passing to the limit yields the result. ¤

Corollary 21.3. The sequence

rF (t)/F ¡ ¢ @ a n−1 ¡ ¢
0 → I n (F ) −−−−→ I n F (t) −→ I F (x) → 0
x∈A1F
is split exact for each n ≥ 1.

¡ ¢
Proof. We show by induction on d = deg x that I n−1 F (x) lies in im(∂). Let
g2 , . . . , gn ∈ F [t] be of degree < d. We need to prove that b = hhḡ2 , . . . , ḡn ii lies in
im(∂) where ḡi is the image of gi in F (x). ` By Example ¡ 19.13,
¢ we have ∂x (c) = b where
n−1
c = hh−fx , g2 , . . . , gn ii. Moreover, c−b ∈ deg x<d I F (x) and therefore c−b ∈ im(∂)
by induction. ¡ ¢
To finish, it suffices to show exactness at I n F (t) . Let b ∈ ker(∂). By Theorem 21.1,
there exists c ∈ W (F ) satisfying rF (t)/F (c) = b.¡ We show
¢ c ∈ I n (F ). Let x ∈ A¡1F be a fixed
¢
rational point and f = t − t(x). ¡Define ¢ ρ : W F (t) → W (F ) by ρ(d) = ∂x hh−f ii · d .
By Lemma 19.14, we have ρ(I n F (t) ⊂ I n (F ) as F (x) = F . By Example 19.13, the
composition ρ ◦ rF (t)/F is the identity. It follows that c = ρ(b) ∈ I n (F ) as needed. ¤
The above was proven by Milnor in [108] for fields of characteristic not two. We wish
to modify the sequence in Theorem 21.1 to the projective line P1F . If x ∈ A1F is of degree
n, let sx : F (x) → F be the F -linear functional
¡ ¢ ¡ ¢
sx tn−1 (x) = 1 and sx ti (x) = 0 for i < n − 1.
The infinite point ∞ corresponds to the 1/t-adic valuation.
¡ ¢ It has residue field F . The
corresponding second residue homomorphism ∂∞ : W F (t) → W (F ) is taken with re-
spect to the prime
¡ ¢ 1/t. So if 0 6= h ∈ F [t] is¡ of ¢degree n and has leading coefficient
a, we have ∂∞ hhi = hai if n is odd and ∂∞ hhi = 0 otherwise. Define (s∞ )∗ to be
−Id : W (F ) → W (F ). The following theorem was first proved by Scharlau in [124] for
fields of characteristic not two.
Theorem 21.4. The sequence
rF (t)/F ¡ ¢ @ a ¡ ¢ s∗
0 → W (F ) −−−−→ W F (t) −→ W F (x) −→ W (F ) → 0
x∈P1F
¡ ¢
is exact with ∂ = (∂x ) and s∗ = (sx )∗ .
Proof. The map (s∞ )∗ is −Id. Hence by Theorem 21.1, it suffices to show s∗ ◦ ∂ is
the zero map.
¡ ¢
As 1-dimensional bilinear forms generate W F (t) , it suffices to check the result on
one-dimensional forms. Let haf1 · · · fn i be a one-dimensional form with fi ∈ F [t] monic
of degree di and a ∈ F × for i = 1, . . . , n. Let xi ∈ A1F satisfy fi = fxi and si = sxi for
1 ≤ i ≤ n. We must show that
X ¡ ¢ ¡ ¢
(sx )∗ ◦ ∂x haf1 · · · fn i = −(s∞ )∗ ◦ ∂∞ haf1 · · · fn i
X∈A1F
in W (F ). Multiplying through by hai, we may also assume that a = 1.

P
Set A = F [t]/(f1 · · · fn ) and d = dim A. Then d = di . Let : F [t] → A be the
canonical epimorphism and set qi = (f1 · · · fn )/fi . We have an F -vector space homomor-
phism
an X
¡ ¢
α: F (xi ) → A given by h1 (xi ), . . . , hn (xi ) 7→ h̄i q̄i for all h ∈ F [t].
i=1
We show that α is an isomorphism. As both spaces have the same dimension,Pit suffices to
show α is monic. As the qi are relatively prime in F [t], we have an equation ni=1 gi qi = 1
with all gi ∈ F [t]. Then the map
a ¡ ¢
A→ F (xi ) given by h̄ → h(x1 )g(x1 ), . . . , h(xn )gn (xn )
¡ ¢
splits α hence α is monic as needed. Set Ai = α F (xi ) for 1 ≤ i ≤ n.
Let s : A → F be the F -linear functional defined by s(t̄d−1 ) = 1 and s(t̄i ) = 0 for
0 ≤ i < d − 1. Define b to be the bilinear form on A over F given by b(f¯, h̄) = s(f¯h̄) for
f, h ∈ F [t]. If i 6= j, we have
¡ ¢
b α(f (xi )), α(h(xj )) = b(f¯q̄i , h̄q̄j ) = s(f¯h̄q̄i q̄j ) = s(0) = 0
for all f, h ∈ F [t]. Consequently, b|Ai is orthogonal to b|Aj if i 6= j.
¡ ¢
Claim: b|Ai ' (si )∗ ∂fi (hf1 · · · fn i) for i = 1, . . . , n.
Let g, h ∈ F [t]. Write
qi gh = c0 + · · · + cdi −1 tdi −1 + fi p
for some ci ∈ F and p ∈ F [t].
By definition, we have
¡ ¢¡ ¢ ¡ ¢
(si )∗ ∂fi (hf1 · · · fn i) g(xi ), h(xi ) = si qi (xi )g(xi )h(xi ) = cdi −1 .
As deg qi = d − di , we have deg qi tdi −1 = d − 1. Thus
¡ ¢
b|Ai α(g(xi )), α(h(xi )) = b(ḡ q̄i , h̄q̄i ) = s(q̄i2 ḡ h̄) = cdi −1 .
and the claim is established.
As ∂f (f1 · · · fn ) = 0 for all irreducible monic polynomials f 6= fi , i = 1, . . . n, in F [t],
we have, by the Claim,
n
X X
¡ ¢ ¡ ¢
b= (si )∗ ∂xi (hf1 · · · fn i) = (sx )∗ ∂x (hf1 · · · fn i)
i=1 x∈A1F
in W (F ).
Suppose that d = 2e is even. The form b is then metabolic as it has a totally isotropic
subspace of dimension e spanned by 1, t̄, . . . , t̄e−1 . We also have (s∞ )∗ ◦ ∂∞ (b) = 0 in this
case.
Suppose that d = 2e + 1. Then b has a totally isotropic subspace spanned by
1, t̄, . . . , t̄e−1 so b ' hai ⊥ c with c metabolic by the Witt Decomposition Theorem 1.28.
¯ }, we see that hai = h1i. As (s∞ )∗ ◦ ∂∞ (b) =
Computing det b on the basis {1, t̄, . . . , td−1
−h1i, the result follows. ¤

Corollary 21.5. Let K ¡be a finite ¢ simple extension of F and s : K → F a non-trivial

F -linear functional. Then s∗ I n (K) ⊂ I n (F ) for all n ≥ 0. Moreover, the induced map
I n (K)/I n+1 (K) → I n (F )/I n+1 (F ) is independent of the non-trivial F -linear functional s
for all n ≥ 0.
Proof.
¡ Let
¢ x lie in A1F with K = F (x). Let b ∈ I n (K). By Lemma 21.3, there exists
c ∈ I n+1 F (t) such that ∂y (c) = 0 for all y ∈ A1F unless y = x in which case ∂x (c) = b.
It follows by Theorem 21.4 that
X
0= (sy )∗ ◦ ∂y (c) = (sx )∗ (b) − ∂∞ (c).
y∈P1F
By Lemma 19.14, we have ∂∞ (c) ∈ I n (F ), so (sx )∗ (b) ∈ I n (F ). Suppose that s : K → F

is another non-trivial F -linear functional. As in the proof of Corollary 20.8, there exists
a c ∈ K × such that (s)∗ (c) = (sx )∗ (cc) for all symmetric bilinear forms c. In particular,
(s)∗ (b) = (sx )∗ (cb) lies in I n (F ). As hhcii · b ∈ I n+1 (K), we also have
¡ ¢
s∗ (b) − (sx )∗ (b) = (sx )∗ hhcii · b
lies in I n+1 (F ). The result follows. ¤
The transfer induced by distinct non-trivial F -linear functionals K → F are not in
general equal on I n (F ).
Exercise 21.6. Show that Corollary 21.5 holds for arbitrary finite extensions K/F .
Corollary 21.7. The sequence
rF (t)/F ¡ ¢ @ a n−1 ¡ ¢ s∗ n−1
0 → I n (F ) −−−−→ I n F (t) −→ I F (x) −→ I (F ) → 0
x∈P1F
is exact.
CHAPTER IV
Function Fields of Quadrics
22. Quadrics
A quadratic form ϕ over F defines a projective quadric Xϕ over F . The quadric Xϕ is

smooth if and only if ϕ is non-degenerate (cf. Proposition 22.1). The quadric Xϕ encodes
information about isotropy properties of ϕ, namely the form ϕ is isotropic over a field
extension E/F if and only if Xϕ has a point over E. In the third part of the book, we
will use algebraic-geometric methods to study isotropy properties of ϕ.
If b is a symmetric bilinear form, the quadric Xϕb reflects isotropy properties of b (and
of ϕb as well). If the characteristic of F is two, only totally singular quadratic forms arise
from symmetric bilinear forms. In particular, quadrics arising from bilinear forms are not
smooth. Therefore, algebraic-geometric methods have wider application in the theory of
quadratic forms than in the theory of bilinear forms.
In the previous sections, we looked at quadratic forms over field extensions determined
by irreducible polynomials. In particular, we were interested in when a quadratic form be-
comes isotropic over such a field. Viewing a quadratic form as a homogeneous polynomial
of degree two, results from these sections apply.
Let ϕ and ψ be two anisotropic quadratic forms. In this section, we begin our study
of when ϕ becomes isotropic or hyperbolic over the function field F (Xψ ) of the integral
quadric Xψ . It is natural at this point to introduce the geometric language that we shall
use, i.e., to associate to a quadratic form a projective quadric.
Let ϕ be a quadratic form on V . Viewing ϕ ∈ S 2 (V ∗ ), we define the projective quadric
associated to ϕ to be the closed subscheme
¡ ¢
Xϕ = Proj S • (V ∗ )/(ϕ)
of the projective space P(V ) = Proj S • (V ∗ ) where S • (V ∗ ) is the symmetric algebra of V ∗ .
The scheme Xϕ is equidimensional of dimension dim V − 2 if ϕ 6= 0 and dim V ≥ 2. We
define the Witt index of Xϕ by i0 (Xϕ ) := i0 (ϕ). By construction, for any field extension
L/F , the set of L-points Xϕ (L) coincides with the set of isotropic lines in VL . Therefore,
Xϕ (L) = ∅ if and only if ϕL is anisotropic.
For any field extension K/F we have XϕK = (Xϕ )K .
Let ψ be a subform of ϕ. The inclusion of vector spaces Vψ ⊂ V gives rise to a
surjective graded ring homomorphism
S • (V ∗ )/(ϕ) → S • (Vψ ∗ )/(ψ)
which in its turn leads to a closed embedding Xψ ,→ Xϕ . We shall always identify Xψ
with a closed subscheme of Xϕ .
91
92 IV. FUNCTION FIELDS OF QUADRICS
Proposition 22.1. Let ϕ be a nonzero quadratic form of dimension at least 2. Then

the quadric Xϕ is smooth if and only if ϕ is non-degenerate.
Proof. By Lemma 7.17, we may assume that F is algebraically closed. We claim
that P(rad ϕ) is the singular locus of Xϕ . Let 0 6= u ∈ V be an isotropic vector. Then the
isotropic line U = F u ⊂ V can be viewed as a rational point of Xϕ . As ϕ(u + εv) = 0 if
and only if u is orthogonal to v (where ε2 = 0), the tangent space TX,U is the subspace
Hom(U, U ⊥ /U ) of the tangent space TP(V ),U = Hom(U, V /U ) (cf. Example 104.20). In
particular, the point U is regular on X if and only if dim TX,U = dim X = dim V − 2 if
and only if U ⊥ 6= V , i.e., U is not contained in rad ϕ. Thus Xϕ is smooth if and only if
rad ϕ = 0, i.e., ϕ is non-degenerate. ¤
We say that the quadratic form ϕ on V is irreducible if ϕ is irreducible in the ring
S (V ∗ ). If ϕ is nonzero and not irreducible, then ϕ = l · l0 for some nonzero linear forms
•
l, l0 ∈ V ∗ . Then rad ϕ = ker(l) ∩ ker(l0 ) has codimension at most 2 in V . Therefore, the

induced form ϕ̄ on V / rad ϕ is either one-dimensional or a hyperbolic plane. It follows
that a regular quadratic form ϕ is irreducible if and only if dim ϕ ≥ 3 or dim ϕ = 2 and
ϕ is anisotropic.
If ϕ is irreducible, Xϕ is an integral scheme. The function field F (Xϕ ) is called the
function field of ϕ and will be denoted by F (ϕ). By definition, F (ϕ) is the subfield of
degree 0 elements in the quotient field of the domain S • (V ∗ )/(ϕ). Note that the quotient
field of S • (V ∗ )/(ϕ) is a purely transcendental extension of F (ϕ) of degree 1. Clearly ϕ is
isotropic over the quotient field of S • (V ∗ )/(ϕ) and therefore is isotropic over F (ϕ).
Example 22.2. Let σ be an anisotropic binary quadratic form. As σ is isotropic over
F (σ), it follows from Corollary 12.3 that F (σ) ' C0 (σ).
If K/F is a field extension such that ϕK is still irreducible, we simply write K(ϕ) for
K(ϕK ).
Example 22.3. Let ϕ and ψ be irreducible quadratic forms. Then F (Xϕ × Xψ ) '
F (ϕ)(ψ) ' F (ψ)(ϕ).
Let ϕ and ψ be two irreducible regular quadratic forms. We shall be interested in when
ϕF (ψ) is hyperbolic or isotropic. A consequence of the Quadratic Similarity Theorem 20.17
is:
Proposition 22.4. Let ϕ be a non-degenerate quadratic form of even dimension and
ψ be an irreducible quadratic form of dimension n over F . Suppose that T = (t1 , . . . , tn )
and b ∈ D(ψ). Then ϕF (ψ) is hyperbolic if and only if
b · ψ(T )ϕF (T ) ' ϕF (T ) .
Proof. By the Quadratic Similarity Theorem 20.17, we have ϕF (ψ) is hyperbolic if
and only if ψ ∗ ψ(T )ϕF (T ) ' ϕF (T ) . Let b ∈ D(ψ). Choosing a basis for V with first vector
v satisfying ψ(v) = b, we have ψ ∗ = b. ¤
Theorem 22.5. ( Subform Theorem) Let ϕ be a nonzero anisotropic quadratic form
and ψ be an irreducible anisotropic quadratic form such that the form ϕF (ψ) is hyperbolic.
22. QUADRICS 93
Let a ∈ D(ϕ) and b ∈ D(ψ). Then abψ is isometric to a subform of ϕ and, therefore,
dim ψ ≤ dim ϕ.
Proof. We view ψ as an irreducible polynomial in F [T ]. The form ϕ is non-degenerate
of even dimension by Remark 7.19, so by Corollary 22.4, we have bψ(T ) ∈ G(ϕF (T ) ). Since
a ∈ D(ϕ), we have abψ(T ) ∈ D(ϕF (T ) ). By the Representation Theorem 17.12, we con-
clude that abψ is a subform of ϕ. ¤
By the proof of the theorem and Corollary 20.21, we have
Corollary 22.6. Let ϕ be an anisotropic quadratic form and ψ an irreducible anisotropic
quadratic form. If ϕF (ψ) is hyperbolic then D(ψ)D(ψ) ⊂ G(ϕ). In particular, if 1 ∈ D(ψ)
then D(ψ) ⊂ G(ϕ).
Remark 22.7. The natural analogues of the Representation Theorem 17.12 and the
Subform Theorem 22.5 are not true for bilinear forms in characteristic two. Let b = h1, bi
and c = h1, ci be anisotropic symmetric bilinear forms with b and c = x2 + by 2 nonzero
and bF ×2 6= cF ×2 in a field F of characteristic two. Thus b 6' c. However, ϕb ' ϕc by
Example 7.28. So ϕc (t1 , t2 ) ∈ D(ϕb F (t1 ,t2 ) ) and cF (ϕb ) is isotropic hence metabolic but ac
is not a subform of b for any a 6= 0.
We do have, however, the following:

Corollary 22.8. Let b and c be anisotropic bilinear forms with dim c ≥ 2 and b
nonzero. Let ψ be the associated quadratic form of c. If bF (ψ) is metabolic then dim c ≤
dim b.
Proof. Let ϕ = ϕb . By the Bilinear Similarity Theorem 20.20 and Lemma 9.3, we
have aψ(T ) ∈ G(bF (T ) ) ⊂ G(ϕF (T ) ) for some a ∈ F × where T = (t1 , . . . , tdim ψ ). It follows
that bψ(T ) ∈ D(ϕF (T ) ) for some b ∈ F × . Consequently,
dim b = dim ϕ ≥ dim ψ = dim c
by the Representation Theorem 17.12. ¤
We turn to the case that a quadratic form becomes isotropic over the function field of
another form or itself.
Proposition 22.9. Let ϕ be an irreducible regular quadratic form. Then the field
extension F (ϕ)/F is purely transcendental if and only if ϕ is isotropic.
Proof. Suppose that the field extension F (ϕ)/F is purely transcendental. As ϕF (ϕ)
is isotropic, ϕ is isotropic by Lemma 7.16.
Now suppose that ϕ is isotropic. Then ϕ = H ⊥ ϕ0 for some ϕ0 by Proposition 7.14.
Let V = Vϕ , V 0 = Vϕ0 and let h, h0 ∈ V be a hyperbolic pair of H. Let ψ = ϕ|F h0 ⊕V 0
with h0 ∈ (V 0 )⊥ . It suffices to show that Xϕ \ Xψ is isomorphic to an affine space. Every
isotropic line in Xϕ \ Xψ has the form F (h + ah0 + v 0 ) for unique a ∈ F and v 0 ∈ V 0
satisfying
0 = ϕ(h + ah0 + v 0 ) = a + ϕ(v 0 ),
i.e., a = −ϕ(v 0 ). Therefore the morphism Xϕ \ Xψ → A(V 0 ) taking F (h + ah0 + v 0 ) to v 0
is an isomorphism with the inverse given by v 0 7→ F (h − ϕ(v 0 )h0 + v 0 ). ¤
Remark 22.10. Let char F = 2 and ϕ an irreducible totally singular form. Then the
field extension F (ϕ)/F is not purely transcendental even if ϕ is isotropic.
Proposition 22.11. Let ϕ be an anisotropic quadratic form over F and K/F a qua-
dratic field extension. Then ϕK is isotropic if and only if there is a binary subform σ of
ϕ such that F (σ) ' K.
Proof. Let σ be a binary subform of ϕ with F (σ) ' K. Since σ is isotropic over
F (σ) we have ϕ isotropic over F (σ) ' K.
Conversely, suppose that ϕK (v) = 0 for some nonzero v ∈ (Vϕ )K . Since K is quadratic
over F , there is a 2-dimensional subspace U ⊂ Vϕ with v ∈ UK . Therefore, the form
σ = ϕ|U is isotropic over K. As σ is also isotropic over F (σ), it follows from Corollary
12.3 and Example 22.2 that F (σ) ' C0 (σ) ' K. ¤
Corollary 22.12. Let ϕ be an anisotropic quadratic form and σ a non-degenerate
anisotropic binary quadratic form. Then ϕ ' b⊗σ ⊥ ψ with b a non-degenerate symmetric
bilinear form and ψF (σ) anisotropic.
Proof. Suppose that ϕF (σ) is isotropic. By Proposition 22.11, there is a binary
subform σ 0 of ϕ with F (σ 0 ) = F (σ). By Corollary 12.2 and Example 22.2, we have σ 0
is similar to σ. Consequently, there exists an a ∈ F × such that ϕ ' aσ ⊥ ψ for some
quadratic form ψ. The result follows by induction on dim ϕ. ¤
Recall that a field extension K/F is called separable if there exists and intermediate
field E in K/F with E/F purely transcendental and K/E algebraic and separable. We
show that regular quadratic forms remain regular after extending to a separable field
extension.
Lemma 22.13. Let ϕ be a regular quadratic form over F and K/F a separable (possibly
infinite) field extension. Then ϕK is regular.
Proof. We proceed in several steps.
Case 1: [K : F ] = 2.
Let v ∈ (Vϕ )K be an isotropic vector. Then v ∈ UK for a 2-dimensional subspace U ⊂ Vϕ
such that ϕ|U is similar to the norm form N of K/F (cf. Proposition 12.1). As N is
non-degenerate, v ∈/ rad bϕK , therefore, rad ϕK = 0.
Case 2: K/F is of odd degree or purely transcendental.
We have ϕ ' ϕan ⊥ nH. The anisotropic part ϕan stays anisotropic over K by Springer’s
Theorem 18.5 or Lemma 7.16 respectively, therefore ϕK is regular.
Case 3: [K : F ] is finite.
We may assume that K/F is Galois by Remark 7.15. Then K/F is a tower of odd degree
and quadratic extensions.
Case 4: The general case.
In general, K/F is a tower of a purely transcendental and a finite separable extension. ¤
We turn to the function field of an irreducible quadratic form.
22. QUADRICS 95
Lemma 22.14. Let ϕ be an irreducible quadratic form over F . Then there exists a
purely transcendental extension E of F with [F (ϕ) : E] = 2. Moreover, if ϕ is not totally
singular, the field E can be chosen with F (ϕ)/E separable. In particular, F (ϕ)/F is
separable.
Proof. Let U ⊂ Vϕ be an anisotropic line. The rational projection f : Xϕ 99K P =
P(V /U ) taking a line U 0 to (U + U 0 )/U is a double cover, so F (ϕ)/E is a quadratic field
extension where E is the purely transcendental extension F (P) of F .
Let τ be the reflection of ϕ with respect to a nonzero vector in U . Clearly, f (τ U 0 ) =
f (U 0 ) for every line U 0 in Xϕ . Therefore, τ induces an automorphism of every fiber of f .
In particular, τ induces an automorphism of the generic fiber hence an automorphism ε
of the field F (ϕ) over E.
If ϕ is not totally singular, we can choose U not in rad bϕ . Then the isometry τ
and the automorphism ε are nontrivial. Consequently, the field extension F (ϕ)/E is
separable. ¤
22.A. Domination relation. Let ϕ and ψ be anisotropic quadratic forms of dimen-
sion at least 2 over F . We say ϕ dominates ψ and write ϕ Â ψ if ϕF (ψ) is isotropic and
write ϕ ≺Â ψ if ϕ Â ψ and ψ Â ϕ. For example, if ψ is a subform of ϕ then ϕ Â ψ.
We have ϕ Â ψ if and only if there exists a rational map Xψ 99K Xϕ .
We show that the relation Â is transitive.
Lemma 22.15. Let ϕ and ψ be anisotropic quadratic forms over F . If ψ Â µ then
there exist a purely transcendental field extension E/F and a binary subform σ of ψE over
E such that E(σ) = F (µ).
Proof. By Lemma 22.14, there exists a purely transcendental field extension E/F
such that F (µ) is a quadratic extension of E. As ψ is isotropic over F (µ), it follows from
Proposition 22.11 applied to the form ψE and the quadratic extension F (µ)/E that ψE
contains a binary subform σ over E satisfying E(σ) = F (µ). ¤
Proposition 22.16. Let ϕ, ψ, and µ be anisotropic quadratic forms over F . If
ϕ Â ψ Â µ then ϕ Â µ.
Proof. Consider first the case when µ is a subform of ψ.
We may assume that µ is of codimension one in ψ. Let T = (t1 , . . . , tn ) be the
coordinates in Vψ so that Vµ is given by t1 = 0. By assumption, there is v ∈ Vϕ [T ] such
that ϕ(v) is divisible by ψ(T ) but v is not divisible by ψ(T ). Since ψ is anisotropic,
we have degti ψ = 2 for every i. Applying the division algorithm on dividing v by ψ
with respect to the variable t2 , we may assume that degt2 v ≤ 1. Moreover, dividing
out a power of t1 if necessary, we may assume that v is not divisible by t1 . Therefore,
the vector w := v|t1 =0 ∈ Vϕ [T 0 ] with T 0 = (t2 , . . . , tn ) is not zero. As degt2 w ≤ 1 and
degt2 µ = 2, the vector w is not divisible by µ(T 0 ). On the other hand, ϕ(w) is divisible
by ψ(T )|t1 =0 = µ(T 0 ), i.e., ϕ is isotropic over F (µ).
Now consider the general case. By Lemma 22.15, there exist a purely transcendental
field extension E/F and a binary subform σ of ψE over E such that E(σ) = F (µ). By
the first part of the proof applied to the forms ϕE Â ψE Â σ, we have ϕE is isotropic over
E(σ) = F (µ), i.e., ϕ Â µ. ¤
Corollary 22.17. Let ϕ, ψ, and µ be anisotropic quadratic forms over F . If
ϕ ≺Â ψ then µF (ϕ) is isotropic if and only if µF (ψ) is isotropic.
Proposition 22.18. Let ψ and µ be anisotropic quadratic forms over F satisfying
ψ Â µ. Let ϕ be a quadratic form such that ϕF (ψ) is hyperbolic. Then ϕF (µ) is hyperbolic.
Proof. Consider first the case when µ is a subform of ψ. Choose variables T 0 of
µ and variables T = (T 0 , T 00 ) of ψ so that µ(T 0 ) = ψ(T 0 , 0). As ϕF (ψ) is hyperbolic,
by the Quadratic Similarity Theorem 20.17, we have ϕF (T ) ' aψ(T )ϕF (T ) over F (T )
for some a ∈ F × . Specializing the variables T 00 = 0, we see by Corollary 20.21 that
ϕF (T 0 ) ' aµ(T 0 )ϕF (T 0 ) over F (T 0 ), and again it follows from the Quadratic Similarity
Theorem 20.17 that ϕF (µ) is hyperbolic.
Now consider the general case. By Lemma 22.15, there exist a purely transcendental
field extension E/F and a binary subform σ of ψE over E such that E(σ) = F (µ). As
ϕE(ψ) is hyperbolic, by the first part of the proof applied to the forms ψE Â σ, we have
ϕE(σ) = ϕF (µ) is hyperbolic. ¤
23. Quadratic Pfister forms II
The introduction of function fields of quadrics allows us to determine the main char-
acterization of general quadratic Pfister forms first proven by Pfister in [111] for fields of
characteristic different from two. They are precisely those forms that become hyperbolic
over their function fields. In particular, Pfister forms can be characterized as universally
round forms.
If ϕ is an anisotropic general quadratic Pfister form then ϕF (ϕ) is isotropic hence
hyperbolic by Corollary 9.11. We wish to show the converse of this property. We begin
by looking at subforms of Pfister forms.
Lemma 23.1. Let ϕ be an anisotropic quadratic form over F and ρ a subform of ϕ.
Suppose that D(ϕK ) and D(ρK ) are groups for all field extensions K/F . Let a = −ϕ(v)
for some v ∈ Vρ⊥ \ Vρ . Then the form hhaii ⊗ ρ is isometric to a subform of ϕ.
Proof. Let T = (t1 , . . . , tn ) and T 0 = (tn+1 , . . . , t2n ) be 2n independent variables
where n = dim ρ. We have
£ ρ(T ) ¤
ρ(T ) − aρ(T 0 ) = ρ(T 0 ) − a .
ρ(T 0 )
ρ(T ) ρ(T )
As D(ρF (T,T 0 ) ) is a group, we have ∈ D(ρF (T,T 0 ) ) hence − a ∈ D(ϕF (T,T 0 ) ). As
ρ(T 0 ) ρ(T 0 )
ρ(T 0 ) ∈ D(ϕF (T,T 0 ) ), we have
ρ(T ) − aρ(T 0 ) ∈ D(ϕF (T,T 0 ) )D(ϕF (T,T 0 ) ) = D(ϕF (T,T 0 ) ).
By the Representation Theorem 17.12, we conclude that hhaii ⊗ ρ is a subform of ϕ. ¤
23. QUADRATIC PFISTER FORMS II 97
Theorem 23.2. Let ϕ be a non-degenerate (respectively, totally singular) n-dimensional

anisotropic quadratic form over F with n ≥ 1. Let T = (t1 , . . . , tn ) and T 0 = (tn+1 , . . . , t2n )
be 2n independent variables. Then the following are equivalent:
(1) n = 2k for some k ≥ 1 and ϕ ∈ Pk (F ) (respectively, ϕ is a quadratic quasi-Pfister
form).
(2) G(ϕK ) = D(ϕK ) for all field extensions K/F .
(3) D(ϕK ) is a group for all field extensions K/F .
(4) Over the rational function field F (T, T 0 ), we have
ϕ(T )ϕ(T 0 ) ∈ D(ϕF (T,T 0 ) ).
(5) ϕ(T ) ∈ G(ϕF (T ) ).
Proof. (2) ⇒ (3) ⇒ (4) are trivial.
(5) ⇐ (1) ⇒ (2): As quadratic Pfister forms are round by Corollary 9.10 and quasi-
Pfister forms are round by Corollary 10.3, the implications follow.
(5) ⇒ (4): We have ϕ(T ) ∈ G(ϕF (T ) ) ⊂ G(ϕF (T,T 0 ) ) and ϕ(T 0 ) ∈ D(ϕT,T 0 ) ). It follows by
Lemma 9.2 that ϕ(T )ϕ(T 0 ) ∈ D(ϕF (T,T 0 ) ).
(4) ⇒ (3): If K/F is a field extension then ϕ(T )ϕ(T 0 ) ∈ D(ϕK(T,T 0 ) ). By the Substitution
Principle 17.7, it follows that D(ϕK ) is a group.
(3) ⇒ (1): As 1 ∈ D(ϕ), it suffices to show that ϕ is a general quadratic (quasi-) Pfister
form. We may assume that dim ϕ ≥ 2. If ϕ is non-degenerate, ϕ contains a non-degenerate
binary subform, i.e., a 1-fold general quadratic Pfister form. Let ρ be the largest quadratic
general Pfister subform of ϕ if ϕ is non-degenerate and the largest quasi-Pfister form if
ϕ is totally singular. Suppose that ρ 6= ϕ. If ϕ is non-degenerate then Vρ⊥ 6= 0 and
Vρ⊥ ∩ Vρ = rad bρ = 0 and if ϕ is totally singular then Vρ⊥ = Vϕ and Vρ 6= Vϕ . In either
case, there exists a v ∈ Vρ⊥ \ Vρ . Set a = −ϕ(v). By Lemma 23.1, we have hhaii ⊗ ρ is
isometric to a subform of ϕ, a contradiction. ¤
Remark 23.3. Let ϕ be a non-degenerate isotropic quadratic form over F . As hyper-
bolic quadratic forms are universal and round, if ϕ is hyperbolic then ϕ(T ) ∈ G(ϕF (T ) ).
Conversely, suppose ϕ(T ) ∈ G(ϕF (T ) ). As
(ϕF (T ) )an ⊥ i0 (ϕ)H ' ϕF (T ) ' ϕ(T )ϕF (T ) ' ϕ(T )(ϕF (T ) )an ⊥ i0 (ϕ)ϕ(T )H,
¡ ¢
we have ϕ(T ) ∈ G (ϕF (T ) )an by Witt Cancellation 8.4. If ϕ was not hyperbolic then the
Subform Theorem 22.5 would imply dim ϕF (T ) ≤ dim(ϕF (T ) )an , a contradiction. Conse-
quently, ϕ(T ) ∈ G(ϕF (T ) ) if and only if ϕ is hyperbolic.
Corollary 23.4. Let ϕ be a non-degenerate anisotropic quadratic form of dimension
at least two over F . Then the following are equivalent:
(1) dim ϕ is even and i1 (ϕ) = dim ϕ/2.
(2) ϕF (ϕ) is hyperbolic.
(3) ϕ ∈ GPn (F ) for some n ≥ 1.
Proof. Statements (1) and (2) are both equivalent to ϕF (ϕ) contains a totally isotropic
subspace of dimension 21 dim ϕ. Let a ∈ D(ϕ). Replacing ϕ by haiϕ we may assume that ϕ
represents one. By Theorem 22.4, condition (2) in the corollary is equivalent to condition
(5) of Theorem 23.2 hence conditions (2) and (3) above are equivalent. ¤
Corollary 23.5. Let ϕ and ψ be quadratic forms over F with ϕ ∈ Pn (F ) anisotropic.
Suppose that there exists an F -isomorphism F (ϕ) ' F (ψ). Then there exists an a ∈ F ×
such that ψ ' aϕ over F , i.e., ϕ and ψ are similar over F .
Proof. As ϕF (ϕ) is hyperbolic so is ϕF (ψ) . In particular, aψ is a subform of ϕ for
some a ∈ F × by the Subform Theorem 22.5. Since F (ϕ) ' F (ψ), we have dim ϕ = dim ψ
and the result follows. ¤
In general the corollary does not generalize to non Pfister forms. Let F = Q(t1 , t2 , t3 )’
The quadratic forms ϕ = hht1 , t2 ii ⊥ h−t3 i and ϕ = hht1 , t3 ii ⊥ h−t2 i have isomorphic
function fields but are not similar. (Cf. [90, Th. XII.2.15].)
Notation 23.6. Let r : F → K be a homomorphism of fields. Denote the kernel of
rK/F : W (F ) → W (K) by W (K/F ) and the kernel of rK/F : Iq (F ) → Iq (K) by Iq (K/F ).
If ϕ is a non-degenerate even dimensional quadratic form over F , we denote by W (F )ϕ
the cyclic W (F )-module in Iq (F ) generated by ϕ.
Corollary 23.7. Let ϕ be an anisotropic quadratic n-fold Pfister form with n ≥ 1 and
ψ an anisotropic quadratic form of even dimension over F . Then there is an isometry
ψ ' b ⊗ ϕ over F for some symmetric bilinear form b over F if and only if ψF (ϕ) is
hyperbolic. In particular, Iq (F (ϕ)/F ) = W (F )ϕ.
Proof. If b is a bilinear form then (b ⊗ ϕ)F (ϕ) = bF (ϕ) ⊗ ϕF (ϕ) is hyperbolic by
Lemma 8.16 as ϕF (ϕ) is hyperbolic by Corollary 9.11. Conversely, suppose that ψF (ϕ) is
hyperbolic. We induct on dim ψ. Assume that dim ψ > 0. By the Subform Theorem 22.5
and Proposition 7.23, we have ψ ' aϕ ⊥ γ for some a ∈ F × and quadratic form γ. The
form γ also satisfies γF (ϕ) is hyperbolic, so the result follows by induction. ¤
23.A. The Hauptsatz. We next prove a fundamental fact about forms in I n (F )
and Iqn (F ) due to Arason and Pfister known as the Hauptsatz and proven in [11].
Theorem 23.8. (Hauptsatz)
(1) Let 0 6= ϕ be an anisotropic quadratic form lying in Iqn (F ). Then dim ϕ ≥ 2n .
(2) Let 0 6= b be an anisotropic bilinear form lying in I n (F ). Then dim(b) ≥ 2n .
Proof. (1): As Iqn (F P)r is additively generated by general quadratic n-fold Pfister
forms, we can write ϕ = 1=i ai ρi in W (F ) for some anisotropic ρi ∈ Pn (F ) and ai ∈ F × .
We prove the result by induction on r. If r = 1 the result is trivial as ρ1 is anisotropic,
so we may assume that r > 1. As (ρr )F (ρr ) is hyperbolic by Corollary 9.11, applying
¡ ¢ P ¡
the restriction
¡ ¢ map rF (ρr )/F : W (F ) → W F (ρr ) to ϕ yields ϕF (ρr ) = r−1 i=1 ai ρi )F (ρr )
in Iqn F (ρ) . If ϕF (ρr ) is hyperbolic then 2n = dim ρ ≤ dim ϕ by the Subform Theorem
22.5. If this does not occur then by induction 2n ≤ dim(ϕF (ρr ) )an ≤ dim ϕ and the result
follows.
(2): PAs I n (F ) is additively generated by bilinear n-fold Pfister forms, we can write
b = r1=i εi ci in W (F ) for some ci anisotropic bilinear n-fold Pfister forms and εi ∈ {±1}.
Let ϕ = ϕcr the quadratic form associated to cr . Then ϕF (ϕ) is isotropic hence (cr )F (ϕ)
24. LINKAGE OF QUADRATIC FORMS 99
is isotropic hence metabolic by Corollary 6.3. If bF (ϕ) is not metabolic then 2n ≤

dim(bF (ϕ) )an ≤ dim b by induction on r. If bF (ϕ) is metabolic then 2n = dim c ≤ dim b by
Corollary 22.8. ¤
An immediate consequence of the Hauptsatz is a solution to a problem of Milnor, viz.,
T T∞ n
Corollary 23.9. ∞ n
i=1 I (F ) = 0 and i=1 Iq (F ) = 0.
The proof of the Hauptsatz for bilinear forms completes the proof of Corollary 6.19
and Theorem 6.20. We have an analogous result for quadratic Pfister forms.
Corollary 23.10. Let ϕ, ψ ∈ GPn (F ). If ϕ ≡ ψ mod Iqn+1 (F ) then ϕ ' aψ for
some a ∈ F × , i.e., ϕ and ψ are similar over F . If, in addition, D(ϕ) ∩ D(ψ) 6= ∅ then
ϕ ' ψ.
Proof. By the Hauptsatz 23.8, we may assume both ϕ and ψ are anisotropic. As
hhaii ⊗ ψ ∈ GPn+1 (F ), we have aψ ≡ ψ mod Iqn+1 (F ) for any a ∈ F × . Choose a ∈ F ×
such that ϕ ⊥ −aψ in Iqn+1 (F ) is isotropic. By the Hauptsatz 23.8, the form ϕ ⊥ −aψ is
hyperbolic hence ϕ = aψ in Iq (F ). As both forms are anisotropic, it follows by dimension
count that ϕ ' aψ by Remark 8.17. If D(ϕ) ∩ D(ψ) 6= ∅ then we can take a = 1. ¤
If ϕ is a nonzero subform of dimension at least two of an anisotropic quadratic form

ρ then ρF (ϕ) is isotropic. As ϕ must also be anisotropic, we have ρ Â ϕ. For general
Pfister forms, we can say more. Let ρ be an anisotropic general quadratic Pfister form.
Then ρF (ρ) is hyperbolic so contains a totally isotropic subspace of dimension (dim ρ)/2.
Suppose that ϕ is a subform of ρ satisfying dim ϕ > (dim ρ)/2. Then ϕF (ρ) is isotropic
hence ϕ Â ρ also. This motivates the following:
Definition 23.11. An anisotropic quadratic form ϕ is called a Pfister neighbor if
there is a general quadratic Pfister form ρ such that ϕ is isometric to a subform of ρ and
dim ϕ > (dim ρ)/2.
For example, non-degenerate anisotropic forms of dimension at most 3 are Pfister
neighbors.
Remark 23.12. Let ϕ be a Pfister neighbor isometric to a subform of a general
quadratic Pfister form ρ with dim ϕ > (dim ρ)/2. By the above, ϕ ≺Â ρ. Let σ be
another general quadratic Pfister form such that ϕ is isometric to a subform of σ and
dim ϕ > (dim σ)/2. As ρ ≺Â ϕ ≺Â σ and D(ρ) ∩ D(σ) 6= ∅, we have σ ' ρ by the
Subform Theorem 22.5. Thus the general Pfister form ρ is uniquely determined by ϕ up
to isomorphism. We call ρ the associated general Pfister form of ϕ. If ϕ represents one
then ρ is a Pfister form.
24. Linkage of quadratic forms
In this section, we look at the quadratic analogue of linkage of bilinear Pfister forms.
The Hauptsatz shows that anisotropic forms in Iqn (F ) have dimension at least 2n . We
shall be interested in those dimensions that are realizable by anisotropic forms in Iqn (F ).
In this section, we determine the possible dimension of anisotropic forms that are the
sum of two general quadratic Pfister forms as well as the meaning of when the sum of
three general n-fold Pfister forms is congruent to zero mod Iqn (F ). We shall return to
and expand these results in §35 and §82.
Proposition 24.1. Let ϕ ∈ GP (F ).
(1) Let ρ ∈ GPn (F ) be a subform of ϕ with n ≥ 1. Then there is a bilinear Pfister
form b such that ϕ ' b ⊗ ρ.
(2) Let b be a general bilinear Pfister form such that ϕb is a subform of ϕ. Then
there is ρ ∈ P (F ) such that ϕ ' b ⊗ ρ.
Proof. We may assume that ϕ is anisotropic of dimension ≥ 2.
(1): Let b be a bilinear Pfister form of the largest dimension such that b⊗ρ is isometric to
a subform ψ of ϕ. As b⊗ρ in non-degenerate, Vψ⊥ ∩Vψ = 0. We claim that ψ = ϕ. Suppose
not. Then Vψ⊥ 6= 0 hence Vψ⊥ \ Vψ 6= ∅. Choose a = −ψ(v) with v ∈ Vψ⊥ \ Vψ . Lemma 23.1
implies that hhaii ⊗ ρ is isometric to a subform of ϕ, contradicting the maximality of b.
(2): We may assume that char F = 2 and b is a Pfister form, so 1 ∈ D(ϕb ) ⊂ D(ϕ). Let
W be a subspace of Vϕ such that ϕ|W ' ϕb . Choose a vector w ∈ W such that ϕb (w) = 1
and write the quasi-Pfister form ϕb = h1i ⊥ ϕ0b where Vϕ0b is any complementary subspace
of F w in Vϕb . Let v ∈ Vϕ satisfy v is orthogonal to Vϕ0b but b(v, w) 6= 0. Then the
restriction of ϕ on W ⊕ F v is isometric to ψ := ϕ0b ⊥ [1, a] for some a ∈ F × . Note that
ψ is isometric to subforms of both of the general Pfister forms ϕ and µ := b ⊗ hha]]. In
particular, ψ and µ are anisotropic. As dim ψ > 21 dim µ, the form ψ is a Pfister neighbor
of µ. Hence ψ ≺Â µ by Remark 23.12. Since ϕF (ψ) is hyperbolic by Proposition 22.18 so
is ϕF (µ) . It follows from the Subform Theorem 22.5 that µ is isomorphic to a subform of
ϕ as 1 ∈ D(µ) ∩ D(ϕ). By the first statement of the proposition, there is a bilinear Pfister
form c such that ϕ ' c ⊗ µ = c ⊗ b ⊗ hha]]. Hence ϕ ' b ⊗ ρ where ρ = c ⊗ hha]]. ¤
Let ρ be a general quadratic Pfister form. We say a general quadratic Pfister form ψ
(respectively, a general bilinear Pfister form b) is a divisor ρ if ρ ' c ⊗ ψ for some bilinear
Pfister form c (respectively, ρ ' b ⊗ µ for some quadratic Pfister form µ). By Proposition
24.1, any general quadratic Pfister subform of ρ is a divisor of ρ and any general bilinear
Pfister form b of ρ whose associated quadratic form is a subform of ρ is a divisor ρ.
Theorem 24.2. Let ϕ1 , ϕ2 ∈ GP (F ) be anisotropic. Let ρ ∈ GP (F ) be a form of
largest dimension such that ρ is isometric to subforms of ϕ1 and ϕ2 . Then
i0 (ϕ1 ⊥ −ϕ2 ) = dim ρ.
Proof. Note that i0 := i0 (ϕ1 ⊥ −ϕ2 ) ≥ d := dim ρ. We may assume that i0 > 1.
We claim that ϕ1 and ϕ2 have isometric non-degenerate binary subforms. To prove the
claim let W be a two-dimensional totally isotropic subspace of Vϕ1 ⊕ V−ϕ2 . As ϕ1 and
ϕ2 are anisotropic, the projections U1 and U2 of W to Vϕ1 and V−ϕ2 = Vϕ2 respectively
are 2-dimensional. Moreover, the binary forms ψ1 := ϕ1 |U1 and ψ2 = ϕ2 |U2 are isometric.
We may assume that ψ1 and ψ2 are degenerate (and therefore, char(F ) = 2). Hence
ψ1 and ψ2 are isometric to ϕb , where b is a 1-fold general bilinear Pfister form. By
Proposition 24.1(2), we have ϕ1 ' b ⊗ ρ1 and ϕ2 ' b ⊗ ρ2 for some ρi ∈ P (F ). Write
ρi = ci ⊗ νi for bilinear Pfister forms ci and 1-fold quadratic Pfister forms νi . Consider
24. LINKAGE OF QUADRATIC FORMS 101
quaternion algebras Q1 and Q2 whose reduced norm forms are similar to b ⊗ ν1 and b ⊗ ν2
respectively. The algebras Q1 and Q2 are split by a quadratic field extension that splits
b. By Theorem 98.20, the algebras Q1 and Q2 have subfields isomorphic to a separable
quadratic extension L/F . By Example 9.8, the reduced norm forms of Q1 and Q2 are
divisible by the non-degenerate norm form of L/F . Hence the forms b ⊗ ν1 and b ⊗ ν2 and
therefore ϕ1 and ϕ2 have isometric non-degenerate binary subforms. The claim is proven.
By the claim, ρ is a general r-fold Pfister form with r ≥ 1. Write ϕ1 = ρ ⊥ ψ1 and
ϕ2 = ρ ⊥ ψ2 for some forms ψ1 and ψ2 . We have ϕ1 ⊥ (−ϕ2 ) ' ψ1 ⊥ (−ψ2 ) ⊥ dH.
Assume that i0 > d. Then the form ψ1 ⊥ (−ψ2 ) is isotropic, i.e., ψ1 and ψ2 have a
common value, say a ∈ F × . By Lemma 23.1, the form hh−aii ⊗ ρ is isometric to subforms
of ϕ1 and ϕ2 , a contradiction. ¤
Corollary 24.3. Let ϕ1 , ϕ2 ∈ GPn (F ) be anisotropic forms. Then the possible values
of i0 (ϕ1 ⊥ −ϕ2 ) are 0, 1, 2, 4, . . . , 2n .
Let ϕ1 ∈ GPm (F ) and ϕ2 ∈ GPn (F ) be anisotropic forms satisfying i(ϕ1 ⊥ −ϕ2 ) =
2r > 0. Let ρ be a general quadratic r-fold Pfister form isometric to a subform of ϕ1 and
to a subform of ϕ2 . We call ρ the linkage of ϕ1 and ϕ2 and say that ϕ1 and ϕ2 are r-linked.
By Proposition 24.1, the linkage ρ is a divisor of ϕ1 and ϕ2 . If m = n and r ≥ n − 1, we
say that ϕ1 and ϕ2 are linked.
Remark 24.4. Let ϕ1 and ϕ2 be general quadratic Pfister forms. Suppose that ϕ1 and
ϕ2 have isometric r-fold quasi-Pfister subforms. Then i0 (ϕ1 ⊥ −ϕ2 ) ≥ 2r and by Theorem
24.2, the forms ϕ1 and ϕ2 have isometric general quadratic r-fold Pfister subforms.
For three n-fold Pfister forms, we have:

Proposition 24.5. Let ϕ1 , ϕ2 , ϕ3 ∈ Pn (F ). If ϕ1 + ϕ2 + ϕ3 ∈ Iqn+1 (F ) then there
exist a quadratic (n − 1)-fold Pfister form ρ and a1 , a2 , a3 ∈ F × such that a1 a2 a3 = 1 and
ϕi ' hhai ii ⊗ ρ for i = 1, 2, 3. In particular, ρ is a common divisor of ϕi for i = 1, 2, 3.
Proof. We may assume that all ϕi are anisotropic Pfister forms by Corollary 9.11. In
addition, we have (ϕ3 )F (ϕ3 ) is hyperbolic. By Proposition 23.10, the form (ϕ1 ⊥ −ϕ2 )F (ϕ3 )
is also hyperbolic. As ϕ3 is anisotropic, ϕ1 ⊥ −ϕ2 cannot be hyperbolic by the Hauptsatz
23.8. Consequently,
(ϕ1 ⊥ −ϕ2 )an ' aϕ3 ⊥ τ
over F for some a ∈ F × and a quadratic form τ by the Subform Theorem 22.5 and
Proposition 7.23. As dim τ < 2n and τ ∈ Iqn+1 (F ), the form τ is hyperbolic by Hauptsatz
23.8 and therefore ϕ1 − ϕ2 = aϕ3 in Iq (F ). It follows that i0 (ϕ1 ⊥ −ϕ2 ) = 2n−1 hence ϕ1
and ϕ2 are linked by Theorem 24.2.
Let ρ be a linkage of ϕ1 and ϕ2 . By Proposition 24.1, we have ϕ1 ' hha1 ii ⊗ ρ
and ϕ2 ' hha2 ii ⊗ ρ for some a1 , a2 ∈ F × . Then ϕ3 is similar to (ϕ1 ⊥ −ϕ2 )an '
−a1 hha1 a2 ii ⊗ ρ, i.e., ϕ3 ' hha1 a2 ii ⊗ ρ. ¤
Corollary 24.6. Let ϕ1 , ϕ2 , ϕ3 ∈ Pn (F ). Suppose that
(24.7) ϕ1 + ϕ2 + ϕ3 ≡ 0 mod Iqn+1 (F ).
Then
en (ϕ1 ) + en (ϕ2 ) + en (ϕ3 ) = 0 in H n (F ).
Proof. By Proposition 24.5, we have ϕi ' hhai ii ⊗ ρ for some ρ ∈ Pn−1 (F ) and
ai ∈ F × for i = 1, 2, 3 satisfying a1 a2 a3 = 1. It follows from Proposition 16.1 that
¡ ¢ ¡ ¢ ¡ ¢
en (ϕ1 ) + en (ϕ2 ) + en (ϕ3 ) = en hha1 ii ⊗ ρ + en hha2 ii ⊗ ρ + en hha3 ii ⊗ ρ
= {a1 a2 a3 }en−1 (ρ) = 0. ¤
25. The submodule Jn (F )

By Corollary 23.4, a general quadratic Pfister form has the following “intrinsic” char-
acterization: a non-degenerate anisotropic quadratic form ϕ of positive even dimension is
a general quadratic Pfister form if and only if the form ϕF (ϕ) is hyperbolic. We shall use
this to characterize elements of Iqn (F ). Let ϕ be a form that is nonzero in Iq (F ). There
exists a field extension K/F such that (ϕK )an is a general quadratic n-fold Pfister form
for some n ≥ 1. The smallest possible such n is called the degree deg(ϕ) of ϕ. We shall
see in Theorem 40.10 that ϕ ∈ Iqn (F ) if and only if deg ϕ ≥ n. In this section, we shall
begin the study of the degree of forms. The ideas in this section are due to Knebusch (cf.
[83] and [84]).
We begin by constructing a tower of field extensions of F with (ϕK )an a general
quadratic n-fold Pfister form where K is the penultimate field K in the tower.
Let ϕ be a non-degenerate quadratic form over F . We construct a tower of fields
F0 ⊂ F1 ⊂ · · · ⊂ Fh and quadratic forms ϕq over Fq for all q = 0, . . . , h as follows: We
start with F0 := F , ϕ0 := ϕan , and inductively set Fq := Fq−1 (ϕq−1 ), ϕq := (ϕFq )an for
q > 0. We stop at Fh such that dim ϕh ≤ 1. The form ϕq is called the qth (anisotropic)
kernel form of ϕ. The tower of the fields Fq is called the generic splitting tower of ϕ.
The integer h is called the height of ϕ and denoted by h(ϕ) . We have h(ϕ) = 0 if and
only if dim ϕan ≤ 1.
Let h = h(ϕ). For any q = 0, . . . , h, the q-th absolute higher Witt index jq (ϕ) of ϕ is
defined as the integer i0 (ϕFq ). Clearly one has
0 ≤ j0 (ϕ) < j1 (ϕ) < · · · < jh (ϕ) = [(dim ϕ)/2].
The set of integers {j0 (ϕ), . . . , jh (ϕ)} is called the splitting pattern of ϕ.
Proposition 25.1. Let ϕ be a non-degenerate quadratic form with h = h(ϕ). The
splitting pattern {j0 (ϕ), . . . , jh (ϕ)} of ϕ coincides with the set of Witt indices i0 (ϕK ) over
all field extensions K/F .
Proof. Let K/F be a field extension. Define a tower of fields K0 ⊂ K1 ⊂ · · · ⊂ Kh
by K0 = K and Kq = Kq−1 (ϕq−1 ) for q > 0. Clearly Fq ⊂ Kq for all q. Let q ≥ 0 be the
smallest integer such that ϕq is anisotropic over Kq . It suffices to show that i0 (ϕK ) = jq (ϕ).
By definition of ϕq and jq , we have ϕFq = ϕq ⊥ jq (ϕ)H. Therefore ϕKq = (ϕq )Kq ⊥
jq (ϕ)H. As ϕq is anisotropic over Kq , we have i0 (ϕKq ) = jq (ϕ).
We claim that the extension Kq /K is purely transcendental. This is clear if q = 0.
Otherwise Kq = Kq−1 (ϕq−1 ) is purely transcendental by Proposition 22.9, since ϕq−1 is
25. THE SUBMODULE Jn (F ) 103
isotropic over Kq−1 by the choice of q and is non-degenerate. It follows from the claim
and Remark 8.9 that i0 (ϕK ) = i0 (ϕKq ) = jq (ϕ). ¤
Corollary 25.2. Let ϕ be a non-degenerate quadratic form over F and K/F be a
purely transcendental extension. Then the splitting patterns of ϕ and ϕK are the same.
Proof. This follows from Lemma 7.16. ¤
We define the relative higher Witt indices iq (ϕ), q = 1, . . . , h(ϕ), of a non-degenerate
quadratic form ϕ to be the differences
iq (ϕ) = jq (ϕ) − jq−1 (ϕ).
Clearly, iq (ϕ) > 0 and iq (ϕ) = ir (ϕs ) for any r > 0 and s ≥ 0 such that r + s = q.
Corollary 25.3. Let ϕ be a non-degenerate anisotropic quadratic form over F of
dimension at least two. Then
i1 (ϕ) = j1 (ϕ) = min{i0 (ϕK ) | K/F a field extension with ϕK isotropic}.
Let ϕ be a non-degenerate non-hyperbolic quadratic form of even dimension over F

with h = h(ϕ). Let F0 ⊂ F1 ⊂ · · · ⊂ Fh be the generic splitting tower of ϕ. The form
ϕh−1 = (ϕFh−1 )an is hyperbolic over its function field hence a general n-fold Pfister form
for some integer n ≥ 1 with ih (ϕ) = 2n−1 by Corollary 23.4. The form ϕh−1 is called the
leading form of ϕ and n is called the degree of ϕ and is denoted by deg ϕ. The field Fh−1
is called the leading field of ϕ. For convenience, we set deg ϕ = ∞ if ϕ is hyperbolic.
Remark 25.4. Let ϕ be a non-degenerate quadratic form of even dimension with the
generic splitting tower F0 ⊂ F1 ⊂ · · · ⊂ Fh . If ϕi = (ϕFi )an with i = 0, . . . , h − 1 then
deg ϕi = deg ϕ.
Notation 25.5. Let ϕ be a non-degenerate quadratic form over F and X = Xϕ . Let
q be an integer satisfying 0 ≤ q ≤ h(ϕ). We shall let Xq := Xϕq and also write jq (X)
(respectively, iq (X)) for jq (ϕ) (respectively, iq (ϕ)).
It is a natural problem to classify non-degenerate quadratic forms over a field F of a

given height. This is still an open problem even for forms of height two. By Corollary
23.4, we do know
Proposition 25.6. Let ϕ be an even dimensional non-degenerate anisotropic qua-
dratic form. Then h(ϕ) = 1 if and only if ϕ ∈ GP (F ).
Proposition 25.7. Let ϕ be a non-degenerate quadratic form of even dimension over
F and K/F a field extension with (ϕK )an an m-fold general Pfister for some m ≥ 1.
Then m ≥ deg ϕ. In particular, deg ϕ is the smallest integer n ≥ 1 such that (ϕK )an is a
general n-fold Pfister form over an extension K/F .
Proof. It follows from Proposition 25.1 that
(dim ϕ − 2m )/2 = i0 (ϕK ) ≤ jh(ϕ)−1 (ϕ) = (dim ϕ − 2deg ϕ )/2,
hence the inequality. ¤
Corollary 25.8. Let ϕ be a non-degenerate quadratic form of even dimension over

F . Then deg ϕE ≥ deg ϕ for any field extension E/F .
For every n ≥ 1 set
Jn (F ) := {ϕ ∈ Iq (F ) | deg ϕ ≥ n} ⊂ Iq (F ).
Clearly J1 (F ) = Iq (F ).
Lemma 25.9. Let ρ ∈ GPn (F ) be anisotropic with n ≥ 1. Let ϕ ∈ Jn+1 (F ). Then
deg(ρ ⊥ ϕ) ≤ n.
Proof. We may assume that ϕ is not hyperbolic. Let ψ = ρ ⊥ ϕ. Let F0 , F1 , . . . , Fh
be the generic splitting tower of ϕ and let ϕi = (ϕFi )an . We show that ρFh is anisotropic.
Suppose not. Choose j maximal such that ρFj is anisotropic. Then ρFj+1 is hyperbolic so
dim ϕj ≤ dim ρ by the Subform Theorem 22.5. Hence
2n = dim ρ ≥ dim ϕj ≥ deg 2deg ϕj = 2deg ϕ ≥ 2n+1
which is impossible. Thus ρFh is anisotropic.
As ϕ is hyperbolic over Fh , we have ψFh ∼ ρFh . Consequently,
deg ψ ≤ deg ψFh = deg ρFh = n
hence deg ψ ≤ n as claimed. ¤
Corollary 25.10. Let ϕ and ψ be even dimensional non-degenerate quadratic forms.
Then deg(ϕ ⊥ ψ) ≥ min(deg ϕ, deg ψ).
Proof. If either ϕ or ψ is hyperbolic, this is trivial, so assume that both forms are
not hyperbolic. We may also assume that ϕ ⊥ ψ is not hyperbolic. Let K/F be a field
extension such that (ϕ ⊥ ψ)K ∼ ρ for some ρ ∈ GPn (K) where n = deg(ϕ ⊥ ψ). Then
ϕK ∼ ρ ⊥ (−ψK ). Suppose that deg ψ > n. Then deg ψK > n and applying the lemma
to the form ρ ⊥ (−ψK ) implies deg ϕK ≤ n. Hence deg ϕ ≤ n = deg(ϕ ⊥ ψ). ¤
Proposition 25.11. Jn (F ) is a W (F )-submodule of Iq (F ) for every n ≥ 1.
Proof. Corollary 25.10 shows that Jn (F ) is a subgroup of Iq (F ). Since deg ϕ =
deg(aϕ) for all a ∈ F × , it follows that Jn (F ) is also closed under multiplication by
elements of W (F ). ¤
Corollary 25.12. Iqn (F ) ⊂ Jn (F ).
Proof. As general quadratic n-fold Pfister forms clearly lie in Jn (F ), the result fol-
lows from Proposition 25.11. ¤
Proposition 25.13. Iq2 (F ) = J2 (F ).
Proof. Let ϕ ∈ J2 (F ) and ϕi = ϕFi with Fi , i = 0, . . . , h, the generic splitting tower.
As deg ϕ ≥ 2, the field Fi is the function field of a smooth quadric of dimension at least
2 over Fi−1 , hence the field Fi−1 is algebraically closed in Fi . Since the form ϕh = 0
has trivial discriminant, by descending induction on i, it follows that ϕ = ϕ0 has trivial
discriminant. By Theorem 13.7 we conclude that ϕ ∈ Iq2 (F ). ¤
26. THE SEPARATION THEOREM 105
Proposition 25.14. J3 (F ) = {ϕ | dim ϕ is even, disc(ϕ) = 1, clif(ϕ) = 1}.

Proof. Let ϕ be an anisotropic form of even dimension and trivial discriminant.
Then ϕ ∈ Iq2 (F ) = J2 (F ) by Theorem 13.7 and Proposition 25.13. Suppose ϕ also has
trivial Clifford invariant. We must show that deg ϕ ≥ 3. Let K be the leading field
of ϕ and ρ its leading form. Then ρ ∈ GPn (F ) with n ≥ 2. Suppose that n = 2. As
e2 (ρ) = 0 in H 2 (K), we have ρ is hyperbolic by Corollary 12.5, a contradiction. Therefore,
ϕ ∈ J3 (F ).
2
Prϕ ∈ J3 (F ). Then ϕ ∈ Iq (F ) by Proposition 25.13. In particular, disc(ϕ) = 1 and
Let
ϕ = i=1 ρi with ρi ∈ GP2 (F ), 1 ≤ i ≤ r. We show that clif(ϕ) = 1 by induction on
P
r. Let ρr = bhha, d]] and K = Fd . Then ϕK ∈ J3 (K) and satisfies ϕK = r−1 i=1 (ρi )K as
(ρr )K is hyperbolic. By induction, clif(ϕK ) = 1. Thus clif(ϕ) lies in kernel of Br(F ) →
Br(K). Therefore the index of clif(ϕ) is at most two. Consequently, clif(ϕ) is represented
by a quaternion algebra, hence there exists a 2-fold quadratic Pfister form σ satisfying
clif(ϕ) = clif(σ). Thus clif(ϕ + σ) = 1 so ϕ + σ lies in J3 (F ) by the first part of the proof.
It follows that σ lies in J3 (F ). Therefore, σ = 0 and clif(ϕ) = 1. ¤
As ē2 is an isomorphism (cf. Theorem 16.3 if char F = 2 and Chapter VIII if char F 6=
2), we have I 3 (F ) = J3 (F ). We shall show that I n (F ) = Jn (F ) for all n in Theorem
40.10. In this section, we show the following due to Arason and Knebusch (cf. [10]).
Proposition 25.15. I m (F )Jn (F ) ⊂ Jn+m (F ).
Proof. Clearly, it suffices to do the case that m = 1. Since 1-fold bilinear Pfister
forms additively generate I(F ), it also suffices to show that if ϕ ∈ Jn (F ) and a ∈ F ×
then hhaii ⊗ ϕ ∈ Jn+1 (F ). Let ψ be the anisotropic part of hhaii ⊗ ϕ. We may assume
that ψ 6= 0.
First suppose that ψ ∈ GP (F ). We prove that deg ψ > n by induction on the height
h of ϕ. If h = 1 then ϕ ∈ GP (F ) and the result is clear. So assume that h > 1. Suppose
that ψF (ϕ) remains anisotropic. Applying the induction hypothesis to the form ϕF (ϕ) , we
have
deg ψ = deg ψF (ϕ) > n.
If ψF (ϕ) is isotropic, it is hyperbolic and therefore dim ψ ≥ dim ϕ by the Subform Theorem
22.5. As h > 1 we have
2deg ψ = dim ψ ≥ dim ϕ > 2deg ϕ ≥ 2n ,

hence deg ψ > n.
Now consider the general case. Let K/F be a field extension such that ψK is Witt
equivalent to a general Pfister form and deg ψK = deg ψ. By the first part of the proof
deg ψ = deg ψK > n. ¤
26. The Separation Theorem
There are anisotropic quadratic forms ϕ and ψ with dim ϕ < dim ψ and ϕF (ψ) isotropic.
For example, this is the case when ϕ and ψ are Pfister neighbors of the same Pfister form.
In this section, we show that if two anisotropic quadratic forms ϕ and ψ are separated by
a power of two, more precisely, if dim ϕ ≤ 2n < dim ψ for some n ≥ 0 then ϕF (ψ) remains
anisotropic.
We shall need the following observation.
Remark 26.1. Let ψ be a quadratic form. Then Vψ contains a (maximal) totally
isotropic subspace of dimension i00 (ψ) := i0 (ψ) + dim(rad ψ). Define the invariant s of a
form by s(ψ) := dim ψ − 2i00 (ψ) = dim ψan − dim(rad ψ). If two quadratic forms ψ and µ
are Witt equivalent then s(ψ) = s(µ).
A field extension L/F is called unirational if there is a field extension L0 /L with L0 /F
purely transcendental. A tower of unirational field extensions is unirational. If L/F is
unirational then every anisotropic quadratic form over F remains anisotropic over L by
Lemma 7.16.
Lemma 26.2. Let ϕ be an anisotropic quadratic form over F satisfying dim ϕ ≤ 2n
for some n ≥ 0. Then there exists a field extension K/F and an (n + 1)-fold anisotropic
quadratic Pfister form ρ over K such that
(1) ϕK is isometric to a subform of ρ.
(2) The field extension K(ρ)/F is unirational.
Proof. Let K0 = F (t1 , . . . , tn+1 ) and let ρ = hht1 , . . . , tn+1 ]]. Then ρ is anisotropic.
Indeed by Corollary 19.6 and induction, it suffices to show hht]] is anisotropic over F (t).
If this is false there is an equation f 2 + f g + tg 2 = 0 with f, g ∈ F [t]. Looking at the
highest term of t in this equation gives either a2 t2n = 0 or b2 t2n+1 = 0 where a, b are the
leading coefficients of f, g respectively. Neither is possible.
Consider the class F of field extensions E/K0 satisfying
(10 ) ρ is anisotropic over E.
(20 ) The field extension E(ρ)/F is unirational.
¡ ¢
We show that K0 ∈ F . By the above ρ is anisotropic. Let L = K0 hh1, t1 ]] . Then
L/F is purely transcendental. As ρL is isotropic, L(ρ)/L is also purely transcendental
and hence so is L(ρ)/F . Since K0 (ρ) ⊂ L(ρ), the field extension K0 (ρ)/F is unirational.
For every field E ∈ F , the form ϕE is anisotropic by (20 ). As ρE is non-degenerate,
the form ρE ⊥ (−ϕE ) is regular. We set
¡ ¢ ¡ ¢
m(E) = i0 ρE ⊥ (−ϕE ) = i00 ρE ⊥ (−ϕE )
and let m be the maximum of the m(E) over all E ∈ F.
Claim 1: We have m(E) ≤ dim ϕ and if m(E) = dim ϕ then ϕE is isometric to a subform
of ρE .
Let W be a totally isotropic subspace in VρE ⊥ V−ϕE of dimension m(E). Since ρE and
ϕE are anisotropic, the projections of W to VρE and V−ϕE = VϕE are injective. This gives
the inequality. Suppose that m(E) = dim ϕ. Then the projection p : W → VϕE is an
isomorphism and the composition
p−1
VϕE −−→ W → VρE
identifies ϕE with a subform of ρE .
26. THE SEPARATION THEOREM 107
Claim 2: m = dim ϕ.
Assume that m < dim ϕ. We derive
¡ a¢ contradiction. Let K ∈ F be a field satisfying
m = m(K) and set τ = (ρK ⊥ −ϕK ) an . As the form ρK ⊥ (−ϕK ) is regular, we have
τ ∼ ρK ⊥ (−ϕK ) and
(26.3) dim ρ + dim ϕ = dim τ + 2m.
Let W be a totally isotropic subspace in VρK ⊥ V−ϕK of dimension m. Let σ denote the
restriction of ρK on VρK ∩ W ⊥ . Thus σ is a subform of ρK of dimension ≥ 2n+1 − m > 2n .
In particular, σ is a Pfister neighbor of ρK . By Lemma 8.10, the natural map VρK ∩W ⊥ →
W ⊥ /W identifies σ with a subform of τ .
We show that condition (20 ) holds for K(τ ). Since σ is a Pfister neighbor of ρK , the
form σ and therefore τ is isotropic over K(ρ). By Lemma 22.14, the extension K(ρ)/K is
separable hence τK(ρ) is regular by Lemma 22.13. Therefore, by Lemma 22.9 the extension
K(ρ)(τ )/K(ρ) is purely transcendental. It follows that K(ρ)(τ ) = K(τ )(ρ) is unirational
over F hence condition (20 ) is satisfied.
¡ ¢
As τ is isotropic over K(τ ), we have m K(τ ) > m, hence K(τ ) ∈ / F . Therefore
condition (10 ) does not hold for K(τ ), i.e., ρK is isotropic and therefore hyperbolic over
K(τ ). As ∅ 6= D(σ) ⊂ D(ρK ) ∩ D(τ ), the form τ is isometric to a subform of ρK by the
Subform Theorem 22.5. Let τ ⊥ be the complementary form of τ in ρK . It follows from
(26.3) that
dim τ ⊥ = dim ρ − dim τ = 2m − dim ϕ < dim ϕ.
As ρK ⊥ (−τ ) ∼ τ ⊥ by Lemma 8.13,
¡
(26.4) τ ⊥ (−τ ) ∼ ρK ⊥ −ϕK ) ⊥ (−τ ) ∼ τ ⊥ ⊥ (−ϕK ).
We now use the invariant s defined in Remark 26.1. Since the space of τ ⊥ (−τ )
contains a totally isotropic subspace of dimension dim τ , it follows from (26.4) and Remark
26.1 that
¡ ¢ ¡ ¢
s τ ⊥ ⊥ (−ϕK ) = s τ ⊥ (−τ ) = 0,
i.e., the form τ ⊥ ⊥ (−ϕK ) contains a totally isotropic subspace of half the dimension of
the form. Since dim ϕ > dim τ ⊥ , this subspace intersects VϕK nontrivially, consequently
ϕK is isotropic contradicting condition (20 ). This establishes the claim.
It follows from the claims that ϕK is isometric to a subform of ρK . ¤
Theorem 26.5. (Separation Theorem) Let ϕ and ψ be two anisotropic quadratic forms
over F . Suppose that dim ϕ ≤ 2n < dim ψ for some n ≥ 0. Then ϕF (ψ) is anisotropic.
Proof. Let ρ be an (n + 1)-fold Pfister form over a field extension K/F as in Lemma
26.2 with ϕK a subform of ρ. By the lemma ψK(ρ) is anisotropic. Suppose that ϕK(ψ) is
isotropic. Then ρK(ψ) is isotropic hence hyperbolic. By the Subform Theorem 22.5, there
exists an a ∈ F such that aψK is a subform of ρ. As dim ψ > 12 dim ρ, the form aψK is a
neighbor of ρ hence aψK(ρ) and therefore ψK(ρ) is isotropic. This is a contradiction. ¤
Corollary 26.6. Let ϕ and ψ be two anisotropic quadratic forms over F with
dim ψ ≥ 2. If dim ψ ≥ 2 dim ϕ − 1 then ϕF (ψ) is anisotropic.
The Separation Theorem was first proved by Hoffmann in [54] for fields of character-
istic different from two and by Hoffmann and Laghribi in [59] for fields of characteristic
two. We shall give another, more geometric proof, in 79.8 below.
27. A further characterization of quadratic Pfister forms

In this section, we give a further characterization of quadratic Pfister forms due to
Fitzgerald (cf. [46]) used to answer a problem of Knebusch in [84] for fields of character-
istic different from two. Knebusch and Scharlau showed in [85] that Fitzgerald’s theorem
implied that if a non-degenerate anisotropic quadratic form ρ becomes hyperbolic over the
function field of an irreducible anisotropic form ϕ satisfying dim ϕ > 13 dim ρ then ρ is a
general quadratic Pfister form. Hoffmann and Laghribi showed that this characterization
was also true for fields of characteristic two in [58].
For a non-degenerate non-hyperbolic quadratic form ρ of even dimension, we set
N (ρ) = dim ρ−2deg ρ . Since the splitting patterns of ρ and ρF (t) are the same by Corollary
25.2, we have N (ρF (t) ) = N (ρ).
Theorem 27.1. Let ρ be a non-hyperbolic quadratic form and ϕ a subform of ρ of
dimension at least 2. Suppose that
(1) ϕ and its complementary form in ρ are anisotropic.
(2) ρF (ϕ) is hyperbolic.
(3) 2 dim ϕ > N (ρ).
Then ρ is an anisotropic general Pfister form.
Proof. Note that ρ is a non-degenerate form of even dimension by Remark 7.19 as
ρF (ϕ) is hyperbolic.
Claim 1: For any field extension K/F with ϕK anisotropic and ρK not hyperbolic, ϕK is
isometric to a subform of (ρK )an .
By Lemma 8.13, the form ρ ⊥ (−ϕ) is Witt equivalent to ψ := ϕ⊥ . In particular,
dim ρ = dim ϕ + dim ψ. Set ρ0 = (ρK )an . It follows from (3) that
¡ ¢
dim ρ0 ⊥ (−ϕK ) ≥ 2deg ρ + dim ϕ > dim ρ − dim ϕ = dim ψ.
As ρ0 ⊥ (−ϕK ) ∼ ψK it follows that the form ρ0 ⊥ (−ϕK ) is isotropic, therefore D(ρ0 ) ∩
D(ϕK ) 6= ∅. Since ρ0K(ϕ) is hyperbolic, the form ϕK is isometric to a subform of ρ0 by the
Subform Theorem 22.5 as needed.
Claim 2: ρ is anisotropic.
Applying Claim 1 to K = F implies that ϕ is isometric to a subform of ρ0 = ρan . Let ψ 0
be the complementary form of ϕ in ρ0 . By Lemma 8.13,
ψ 0 ∼ ρ0 ⊥ (−ϕ) ∼ ρ ⊥ (−ϕ) ∼ ψ.
As both forms ψ and ψ 0 are anisotropic, we have ψ 0 ' ψ. Hence
dim ρ = dim ϕ + dim ψ = dim ϕ + dim ψ 0 = dim ρ0 = dim ρan .
Therefore ρ is anisotropic.
27. A FURTHER CHARACTERIZATION OF QUADRATIC PFISTER FORMS 109
We now investigate the form ϕF (ρ) . Suppose it is isotropic. Then ϕ ≺Â ρ hence ρF (ρ)
is hyperbolic by Proposition 22.18. It follows that ρ is a general Pfister form by Corollary
23.4 and we are done. Thus we may assume that ϕF (ρ) is anisotropic. Normalizing we
may also assume that 1 ∈ D(ϕ). We shall prove that ρ is a Pfister form by induction on
dim ρ. Suppose that ρ is not a Pfister form. In particular, ρ1 := (ρF (ρ) )an is nonzero and
dim ρ1 ≥ 2. We shall finish the proof by obtaining a contradiction. Let ϕ1 = ϕF (ρ) .
Note that deg ρ1 = deg ρ and dim ρ1 < dim ρ hence N (ρ1 ) < N (ρ).
Claim 3: ρ1 is a Pfister form.
Applying Claim 1 to the field K = F (ρ), we see that ϕ1 is isometric to a subform of ρ1 .
We have
2 dim ϕ1 = 2 dim ϕ > N (ρ) > N (ρ1 ).
By the induction hypothesis applied to the form ρ1 and its subform, ϕ1 , we conclude that
the form ρ1 is a Pfister form proving the claim. In particular, dim ρ1 = 2deg ρ1 = 2deg ρ .
Claim 4: D(ρ) = G(ρ).
Since G(ρ) ⊂ D(ρ), it suffices to show if x ∈ D(ρ) then x ∈ G(ρ). Suppose that x ∈ / G(ρ).
Hence the anisotropic part β of the isotropic form hhxii ⊗ ρ is nonzero. It follows from
Proposition 25.15 that deg β ≥ 1 + deg ρ.
Suppose that βF (ρ) is hyperbolic. As ρ − β = −xρ in Iq (F ), the form ρ ⊥ (−β) is
isotropic, hence D(ρ) ∩ D(β) 6= ∅. It follows from this that ρ is isometric to a subform of
β by the Subform Theorem 22.5. Let β ' ρ ⊥ µ ∼ ρ ⊥ (−xρ) for some form µ. By Witt
cancellation, µ ∼ −xρ. But dim β < 2 dim ρ hence dim µ < dim ρ. As ρ is anisotropic,
this is a contradiction. It follows that the form β1 = (βF (ρ) )an is not zero and hence
dim β1 ≥ 2deg β ≥ 21+deg ρ .
Since ρ is hyperbolic over F (ϕ), it follows from the Subform Theorem 22.5 that ϕ is
isometric to a subform of xρ. Applying Claim 1 to the form xρF (ρ) , we conclude that
ϕ1 is a subform of xρ1 . As ϕ1 is also a subform of ρ1 , the form hhxii ⊗ ρ1 contains
ϕ1 ⊥ (−ϕ1 ) and¡ therefore ¢ a totally isotropic subspace of dimension dim ϕ1 = dim ϕ.
Therefore dim hhxii ⊗ ρ1 an ≤ 2 dim ρ1 − 2 dim ϕ. Consequently,
¡ ¢
21+deg ρ ≤ dim β1 = dim hhxii ⊗ ρ1 an ≤ 2 dim ρ1 − 2 dim ϕ < 21+deg ρ ,
a contradiction. This proves the claim.
Let F (T ) = F (t1 , . . . , tn ) with n = dim ρ. We have deg ρF (T ) = deg ρ and N (ρF (T ) ) =
N (ρ). Working over F (T ) instead of F , we have the forms ϕF (T ) and ρF (T ) satisfy the
conditions of the theorem. By Claim 4, we conclude that G(ρF (T ) ) = D(ρF (T ) ). It follows
from Theorem 23.2 that ρ is a Pfister form, a contradiction. ¤
Corollary 27.2. Let ρ be a nonzero anisotropic quadratic form and ϕ an irreducible
anisotropic quadratic form satisfying dim ϕ > 31 dim ρ. If ρF (ϕ) is hyperbolic then ρ ∈
GP (F ).
Proof. As ρF (ϕ) is hyperbolic, the form ρ is non-degenerate. It follows by the Subform
Theorem 22.5 that aϕ is a subform of ρ for some a ∈ F × . As ρ is anisotropic, the
complementary form of aϕ in ρ is anisotropic.
Let K be the leading field of ρ and τ its leading form. We show that ϕK is anisotropic.
If ϕF (ρ) is isotropic then ϕ ≺Â ρ. In particular, ρF (ρ) is hyperbolic by Proposition 22.18
hence K = F and ϕ is anisotropic by hypothesis. Thus we may assume that ϕF (ρ)
is anisotropic. The assertion now follows by induction on h(ρ). As τK(ϕ) ∼ ρK(ϕ) is
hyperbolic, dim ϕ = dim ϕK ≤ dim τ = 2deg ρ by the Subform Theorem 22.5. Hence
N (ρ) = dim ρ − 2deg ρ ≤ dim ρ − dim ϕ < 2 dim ϕ. The result follows by Theorem 27.1. ¤
The restriction dim ϕ > 13 dim ρ above is best possible (cf. [85]).
A further application of Theorem 27.1 is given by:
Theorem 27.3. Let ϕ and ψ be non-degenerate quadratic forms over F of the same
odd dimension. If i0 (ϕK ) = i0 (ψK ) for any field extension K/F then ϕ and ψ are similar.
Proof. We may assume that ϕ and ψ are anisotropic and have the same determinants
(cf. Remark 13.8). Let n = dim ϕ. We shall show that ϕ ' ψ by induction on n. The
statement is obvious if n = 1, so assume that n > 1.
We construct a non-degenerate form ρ of dimension 2n and trivial discriminant con-
taining ϕ such that ϕ⊥ ' −ψ as follows: If char F 6= 2 let ρ = ϕ ⊥ (−ψ). If char F = 2
write ϕ ' hai ⊥ ϕ0 and ψ ' hai ⊥ ψ 0 for some a ∈ F × and non-degenerate forms ϕ0 and
ψ 0 . Set ρ = [a, c] ⊥ ϕ0 ⊥ ψ 0 , where c is chosen so that disc ρ is trivial.
By induction applied to the anisotropic parts of ϕF (ϕ) and ψF (ϕ) , we have ϕF (ϕ) '
ψF (ϕ) . It follows from Witt Cancellation and Proposition 13.6 (in the case char F = 2)
that ρF (ϕ) is hyperbolic. If ρ itself is not hyperbolic, then by Theorem 27.1, the form ρ
is an anisotropic general Pfister form of dimension 2n. In particular, n is a power of 2, a
contradiction.
Thus ρ is hyperbolic. By Lemma 8.13, we have −ϕ ∼ ρ ⊥ (−ϕ) ∼ ϕ⊥ ' −ψ. As ϕ
and ψ have the same dimension we conclude that ϕ ' ψ. ¤
28. Excellent quadratic forms
In general, if ϕ is a non-degenerate quadratic form and K/F a field extension then

the anisotropic part of ϕK will not necessarily be isometric to a form defined over F and
extended to K. Those forms over a field F whose anisotropic part over any extension
field of F is defined over F are called excellent forms. These forms were first introduced
by Knebusch in [84]. We study them in this section.
Let K/F be a field extension and ψ a quadratic form over K. We say that ψ is defined
over F if there is a quadratic form η over F such that ψ ' ηK .
Theorem 28.1. Let ϕ be an anisotropic non-degenerate quadratic form of dimension
≥ 2. Then ϕ is a Pfister neighbor if and only if the quadratic form (ϕF (ϕ) )an is defined
over F .
Proof. Let ϕ be a Pfister neighbor and let ρ be the associated general Pfister form
so ϕ is a subform of ρ. As ϕF (ϕ) is isotropic, the general Pfister form ρF (ϕ) is hyperbolic
by Corollary 9.11. By Lemma 8.13, the form ϕF (ϕ) is Witt equivalent to −(ϕ⊥ )F (ϕ) . Since
dim ϕ⊥ < (dim ρ)/2, it follows by Corollary 26.6 that (ϕ⊥ )F (ρ) is anisotropic. By Corollary
28. EXCELLENT QUADRATIC FORMS 111
22.17, the form (ϕ⊥ )F (ϕ) is also anisotropic as ϕ ≺Â ρ by Remark 23.12. Consequently,
(ϕF (ϕ) )an ' (−ϕ⊥ )F (ϕ) is defined over F .
Suppose now that (ϕF (ϕ) )an ' ψF (ϕ) for some (anisotropic) form ψ over F . Note that
dim ψ < dim ϕ.
Claim: There exists a form ρ satisfying
(1) ϕ is a subform of ρ.
(2) The complementary form ϕ⊥ is isometric to −ψ.
(3) ρF (ϕ) is hyperbolic.
Moreover, if dim ϕ ≥ 3, then ρ can be chosen in Iq2 (F ).
Suppose that dim ϕ is even or char F 6= 2. Then ρ = ϕ ⊥ (−ψ) satisfies (1), (2), and
(3). As F is algebraically closed in F (ϕ), if dim ϕ ≥ 3, we have disc ϕ = disc ψ hence
ρ ∈ Iq2 (F ).
So we may assume that char F = 2 and dim ϕ is odd. Write ϕ = ϕ0 ⊥ hai and
ψ = ψ 0 ⊥ hbi for non-degenerate forms ϕ0 , ψ 0 and a, b ∈ F × . Note that hai (respectively,
hbi) is the restriction of ϕ (respectively, ψ) on rad bϕ (respectively, rad bψ ) by Proposition
7.32. By definition of ψ we have haiF (ϕ) ' hbiF (ϕ) . Since F (ϕ)/F is a separable field
extension by Lemma 22.14, we have hai ' hbi. Therefore, we may assume that b = a.
Choose c ∈ F such that disc(ϕ0 ⊥ ψ 0 ) = disc[a, c] and set ρ = ϕ0 ⊥ ψ 0 ⊥ [a, c] so
that ρ ∈ Iq2 (F ). Clearly ϕ is a subform of ρ and ϕ⊥ is isometric to ψ. By Lemma 8.13,
ρ ⊥ ϕ ∼ ψ. Since ϕ and ψ are Witt equivalent over F (ϕ), we have ρF (ϕ) ⊥ ϕF (ϕ) ∼ ϕF (ϕ) .
Cancelling the non-degenerate form ϕ0F (ϕ) yields
ρF (ϕ) ⊥ haiF (ϕ) ∼ haiF (ϕ) .
As ρ ∈ Iq2 (F ) by Proposition 13.6, we have ρF (ϕ) ∼ 0 establishing the claim.
As dim ρ = dim ϕ + dim ψ < 2 dim ϕ and ϕ is anisotropic, it follows that ρ is not
hyperbolic. Moreover, ϕ and its complement ϕ⊥ ' −ψ are anisotropic. Consequently, ρ
is a general Pfister form by Theorem 27.1 hence ϕ is a Pfister neighbor. ¤
Exercise 28.2. Let ϕ be a non-degenerate quadratic form of odd dimension. Then
h(ϕ) = 1 if and only if ϕ is a Pfister neighbor of dimension 2n − 1 for some n ≥ 1.
Theorem 28.3. Let ϕ be a non-degenerate quadratic form. Then the following two
(1) For any field extension K/F , the form (ϕK )an is defined over F .
(2) There are anisotropic Pfister neighbors ϕ0 = ϕan , ϕ1 , . . . , ϕr with associated gen-
eral Pfister forms ρ0 , ρ1 , . . . , ρr respectively satisfying ϕi ' (ρi ⊥ ϕi+1 )an for all
i = 0, 1, . . . , r (with ϕr+1 := 0).
Proof. (2) ⇒ (1): Let K/F be a field extension. If all general Pfister forms ρi are
hyperbolic over K, the isomorphisms in (2) show that all the ϕi are also hyperbolic. In
particular, (ϕK )an is the zero form and hence is defined over F .
Let s be the smallest integer such that (ρs )K is not hyperbolic. Then the forms ϕ =
ϕ0 , ϕ1 , . . . , ϕs are Witt equivalent and (ϕs )K is a Pfister neighbor of the anisotropic general
Pfister form (ρs )K . In particular (ϕs )K is anisotropic and therefore (ϕK )an = (ϕs )K is
defined over F .
(1) ⇒ (2): We prove the statement by induction on dim ϕ. We may assume that dim ϕan ≥
2. By Theorem 28.1, the form ϕan is a Pfister neighbor. Let ρ be the associated general
Pfister form of ϕan . Consider the negative of the complimentary form ψ = −(ϕan )⊥ of
ϕan in ρ. It follows from Lemma 8.13 that ϕan ' (ρ ⊥ ψ)an .
We claim that the form ψ satisfies (1). Let K/F be a field extension. If ρ is hyperbolic
over K then ϕK and ψK are Witt equivalent. Therefore (ψK )an ' (ϕK )an is defined
over F . If ρK is anisotropic then so is ψK , therefore (ψK )an = ψK is defined over F .
By the induction hypothesis applied to ψ, there are anisotropic Pfister neighbors ϕ1 =
ψ, ϕ2 , . . . , ϕr with the associated general Pfister forms ρ1 , ρ2 , . . . , ρr respectively such that
ϕi ' (ρi ⊥ ϕi+1 )an for all i = 1, . . . , r, where ϕr+1 = 0. To finish the proof let ϕ0 = (ϕ)an
and ρ0 = ρ. ¤
A quadratic form ϕ satisfying equivalent conditions of Theorem 28.3 is called excellent.
By Lemma 8.13, the form ϕi+1 in Theorem 28.3(2) is isometric to the negative of the
complement of ϕi in ρi . In particular, the sequences of forms ϕi and ρi are uniquely
determined by ϕ up to isometry. Note that all forms ϕi are also excellent – this allows
inductive proofs while working with excellent forms.
Example 28.4. If char F 6= 2 then the form nh1i is excellent for every n > 0.
Proposition 28.5. Let ϕ be an excellent quadratic form. Then in the notation of
Theorem 28.3 we have the following:
(1) The integer r coincides with the height of ϕ.
(2) If F0 = F, F1 , . . . , Fr is the generic splitting tower of ϕ then (ϕFi )an ' (ϕi )Fi for
all i = 0, . . . , r.
Proof. The last statement is obvious if i = 0. As ρ0 is hyperbolic over F1 = F (ϕan ) =
F (ϕ0 ), the forms ϕF1 and (ϕ1 )F1 are Witt equivalent. Since dim ϕ1 < (dim ρ0 )/2, the
form ϕ1 is anisotropic over F (ρ0 ) by Corollary 26.6. As ϕ0 ≺Â ρ0 , the form ϕ1 is also
anisotropic over F1 = F (ϕ0 ) by Corollary 22.17. Therefore, (ϕF1 )an ' (ϕ1 )F1 . This proves
the last statement for i = 1. Both statements of the proposition follow now by induction
on r. ¤
The notion of an excellent form was introduced by Knebusch.
29. Excellent field extensions
A field extension E/F is called excellent if the anisotropic part ϕE of any quadratic
form ϕ over F is defined over F , i.e., there is a quadratic form ψ over F satisfying
(ϕE )an ' ψE . The concept of an excellent field extension was introduced in [40] and
further information about this concept can be found there.
Example 29.1. Suppose that every anisotropic form over F remains anisotropic over
E. Then for every quadratic form ϕ over F the form (ϕan )E is anisotropic and therefore
is isometric to the anisotropic part of ϕE . It follows that E/F is an excellent field
extension. In particular, it follows from Lemma 7.16 and Springer’s Theorem 18.5 that
purely transcendental field extensions and odd degree field extensions are excellent.
29. EXCELLENT FIELD EXTENSIONS 113
Example 29.2. Let E/F be a separable quadratic field extension. Then E = F (σ),
where σ is the (non-degenerate) binary norm form of E/F . It follows from Corollary
22.12 that E/F is an excellent field extension.
Example 29.3. Let E/F be a field extension such that every quadratic form over E
is defined over F . Then E/F is obviously an excellent extension.
Exercise 29.4. Let E be either an algebraic closure or a separable closure of a field
F . Prove that every quadratic form over E is defined over F . In particular E/F is an
excellent extension.
Let ρ be an irreducible non-degenerate quadratic form over F . If dim ρ = 2, the
extension F (ρ)/F is separable quadratic and therefore is excellent by Example 29.2. We
extend this result to non-degenerate forms of dimension 3.
Notation 29.5. Until the end of this section, let K/F be a separable quadratic field
extension and let a ∈ F × . Consider the 3-dimensional quadratic form ρ = NK/F ⊥ h−ai
on the space U := K ⊕ F . Let X be the projective quadric of ρ. It is a smooth conic
curve in P(U ). In the projective coordinates [s : t] on K ⊕ F , the conic X is given by the
equation NK/F (s) = at2 . We write E for the field F (ρ) = F (X).
The intersection of X with P(K) is Spec F (x) for a point x ∈ X of degree 2 with
F (x) ' K. In fact, Spec F (x) is the quadric of the form NK/F = ρ|K . Over K the norm
form NK/F (s) factors into a product s · s0 of linear forms. Therefore, there are two rational
points y and y 0 of the curve XK mapping to x under the natural morphism XK → X, so
that the divisor div(s/t) equals y − y 0 and div(s0 /t) equals y 0 − y. Moreover, we have
(29.6) NKE/E (s/t) = NK/F (s)/t2 = at2 /t2 = a.
For any n ≥ 0, let Ln be the F -subspace
{f ∈ E × | div(f ) + nx ≥ 0} ∪ {0}
of E. We have
F = L0 ⊂ L1 ⊂ L2 ⊂ · · · ⊂ E
and Ln · Lm ⊂ Ln+m for all n, m ≥ 0. In particular, the union L of all Ln is a subring of
E. In fact, E is the quotient field of L.
In addition, OX,x · Ln ⊂ Ln and mX,x · Ln ⊂ Ln−1 for every n ≥ 1 where OX,x is
the local ring of x and mX,x its maximal ideal. In particular, we have the structure of a
K-vector space on Ln /Ln−1 for every n ≥ 1.
Set Ln = Ln /Ln−1 for n ≥ 1 and L0 = K. The graded group L∗ has the structure of
a ring.
The following lemma is an easy case of the Riemann-Roch Theorem.
Lemma 29.7. In the notation above, we have dimK (Ln ) = 1 for all n ≥ 0. Moreover,
L∗ is a polynomial ring over K in one variable.
Proof. Let f, g ∈ Ln \ Ln−1 for n ≥ 1. Since f = (f /g)g and f /g ∈ (OX,x )× , the

images of f and g in Ln are linearly dependent over K. Hence dimK (Ln ) ≤ 1. On the
other hand, for a nonzero linear form l on K, we have div(l/t) = z − x for some z 6= x.
Hence (l/t)n ∈ Ln \ Ln−1 and therefore dimK (Ln ) ≥ 1. Moreover, L∗ = K[l/t]. ¤
Proposition 29.8. Let ϕ : V → F be an anisotropic quadratic form and suppose that
for some n ≥ 1 there exists
v ∈ (V ⊗ Ln ) \ (V ⊗ Ln−1 )
such that ϕ(v) = 0. Then there exists a subspace W ⊂ V of dimension 2 satisfying
(1) ϕ|W is similar to NK/F .
(2) There exists a nonzero ṽ ∈ V ⊗ Ln−1 such that ϕ̃(ṽ) = 0 where ϕ̃ is the quadratic
form a(ϕ|W ) ⊥ ϕ|W ⊥ on V .
Proof. Let v̄ denote the image of v under the canonical map V ⊗ Ln → V ⊗ Ln . We
have v̄ 6= 0 as v ∈
/ V ⊗ Ln−1 . Since Ln is 2-dimensional over F by Lemma 29.7, there is a
subspace W ⊂ V of dimension 2 with v̄ ∈ W ⊗ Ln .
As v̄ is an isotropic vector in W ⊗ L∗ and L∗ is a polynomial algebra over K, we have
W ⊗ K is isotropic. It follows from Corollary 22.12 that the restriction ϕ|W is isometric
to c NK/F for some c ∈ F × and, in particular, non-degenerate.
By Proposition 7.23, we can write v = w +w0 with w ∈ W ⊗Ln and w0 ∈ W ⊥ ⊗Ln . By
construction of W , we have w̄0 = 0 in V ⊗Ln , i.e., w0 ∈ V ⊗Ln−1 , therefore ϕ(w0 ) ∈ L2n−2 .
Since 0 = ϕ(v) = ϕ(w) + ϕ(w0 ), we must have ϕ(w) ∈ L2n−2 .
We may therefore assume that W = K and ϕ|K = c NK/F .
Thus we have w ∈ K ⊗ Ln ⊂ K ⊗ E = K(X). Considering w as a function on XK
we have div∞ (w) = my + m0 y 0 for some m, m0 ≤ n where div∞ is the divisor of poles. As
w∈/ W ⊗ Ln−1 , we must have one of the numbers m and m0 , say m, equal to n.
Let σ be the generator of the Galois group of K/F . We have σ(y) = y 0 , hence
div∞ (σw) = my 0 + m0 y and
¡ ¢ ¡ ¢
div∞ ϕ(w) = div∞ NK/F (w) = div∞ (w) + div∞ (σw) = (m + m0 )(y + y 0 ).
As ϕ(w) ∈ L2n−2 , we have m + m0 ≤ 2n − 2, i.e., m0 ≤ n − 2.
Note also that
div∞ (ws/t) = div∞ (w) + y − y 0 = (m − 1)y + (m0 + 1)y 0 .
As both m − 1 and m0 + 1 are at most n − 1, we have ws/t ∈ K ⊗ Ln−1 .
Now let ϕ̃ be the quadratic form a(ϕ|W ) ⊥ ϕ|W ⊥ on V = W ⊕ W ⊥ and set ṽ =
a ws/t + w0 ∈ V ⊗ Ln−1 . By (29.6) we have
−1
ϕ̃(ṽ) = aϕ(a−1 ws/t) + ϕ(w0 ) = a−1 NK(X)/F (X) (s/t)ϕ(w) + ϕ(w0 ) = ϕ(w) + ϕ(w0 ) = 0. ¤
Corollary 29.9. Let ϕ be a quadratic form over F such that ϕE is isotropic. Then
there exists an isotropic quadratic form ψ over F such that ψE ' ϕE .
30. CENTRAL SIMPLE ALGEBRAS OVER FUNCTION FIELDS OF QUADRATIC FORMS 115
Proof. Let v ∈ V ⊗ E be an isotropic vector of ϕE . Scaling v we may assume that

v ∈ V ⊗ L. Choose the smallest n such that v ∈ V ⊗ Ln . We induct on n. If n = 0, i.e.,
v ∈ V , the form ϕ is isotropic and we can take ψ = ϕ.
Suppose that n ≥ 1. By Proposition 29.8, there exist a 2-dimensional subspace W ⊂ V
such that ϕ|W is similar to NK/F and an isotropic vector ṽ ∈ V ⊗ Ln−1 for the quadratic
form ϕ̃ = a(ϕ|W ) ⊥ (ϕ|W ⊥ ) on V . As a is the norm in the quadratic extension KE/E,
the forms NK/F and aNK/F are isometric over E, hence ϕ̃E ' ϕE . By the induction
hypothesis applied to the form ϕ̃, there is an isotropic quadratic form ψ over F such that
ψE ' ϕ̃E ' ϕE . ¤
Theorem 29.10. Let ρ be a non-degenerate 3-dimensional quadratic form over F .
Then the field extension F (ρ)/F is excellent.
Proof. We may assume ρ is the form in Notation 29.5 as every non-degenerate 3-
dimensional quadratic form over F is similar to such a form. Let E = F (ρ) and let ϕ be
a quadratic form over F . We show by induction on dim ϕan that (ϕE )an is defined over
F . If ϕan is anisotropic over E we are done as (ϕE )an ' (ϕan )E .
Suppose that ϕan is isotropic over E. By Corollary 29.9 applied to ϕan , there exists an
isotropic quadratic form ψ over F such that ψE ' (ϕan )E . As dim ψan < dim ϕan , by the
induction hypothesis, there is a quadratic form µ over F such that (ψE )an ' µE . Since
µE ∼ ψE ∼ ϕE , we have (ϕE )an ' µE . ¤
Corollary 29.11. Let ϕ ∈ GP2 (F ). Then F (ϕ)/F is excellent.
Proof. Let ψ be a Pfister neighbor of ϕ of dimension three. Let K = F (ϕ) and
L = F (ψ). By Remark 23.12 and Proposition 22.9, the field extensions KL/K and
KL/L are purely transcendental. Let ν be a quadratic form over F . By Theorem 29.10,
there exists a quadratic form σ over F such that (νL )an ' σL . Hence
¡ ¢ ¡ ¢
(νK )an KL ' (νKL )an ' (νL )an KL ' σKL .
It follows that (νK )an ' σK . ¤
The corollary above was first proven by Arason (cf. [40, Appendix II]). We have
followed here Rost’s approach in [121]. This result does not generalize to higher fold
Pfister forms. It is known, in general, for every n > 2, there exists a field F and a
ϕ ∈ GPn (F ) with F (ϕ)/F not an excellent extension (cf. [63]).
30. Central simple algebras over function fields of quadratic forms
Let D be a finite dimensional division algebra over a field F . Let D[t] denote the
F [t]-algebra D ⊗F F [t] and D(t) denote the F (t)-algebra D ⊗F F (t). As D(t) has no zero
divisors and is finite dimensional over F (t), it is a division algebra. The main result in
this section is Theorem 30.5. This was originally proved in [104] for fields of characteristic
different from two using Swan’s calculation of Grothendieck group of a smooth projective
quadric. We generalize the elementary proof for this case given by Tignol in [136].
A subring A ⊂ D(t) is called an order over F [t] if it is a finitely generated F [t]-
submodule of D(t).
Lemma 30.1. Let D be a finite dimensional division F -algebra. Then every order
A ⊂ D(t) over F [t] is conjugate to a subring of D[t].
Proof. As A is finitely generated as an F [t]-module, there is a nonzero f ∈ F [t] such
that Af ⊂ D[t]. The subset DAf of D[t] is a left ideal. The ring D[t] admits both the left
and the right Euclidean algorithm relative to degree. In follows that all one-sided ideals
in D[t] are principal. In particular DAf = D[t]x for some x ∈ D[t]. As A is a ring, for
every y ∈ A, we have
xy ∈ D[t]xy = DAf y ⊂ DAf = D[t]x,
hence xyx−1 ∈ D[t]. Thus xAx−1 ⊂ D[t]. ¤
Lemma 30.2. Suppose that R is a commutative ring and S a (not necessarily com-
mutative) R-algebra. Let X ⊂ S be an R-submodule generated by n elements. Suppose
that every x ∈ X satisfies the equation x2 + ax + b = 0 for some a, b ∈ R. Then the
R-subalgebra of S generated by X can be generated as an R-module by 2n elements.
Proof. Let x1 , . . . , xn be generators for the R-module X. Writing quadratic equa-
tions for every pair of generators xi , xj and xi +xj , we see that xi xj +xj xi +axi +bxj +c = 0
for some a, b, c ∈ R. Therefore, the R-subalgebra of S generated by X is generated as an
R-module by the monomials xi1 xi2 . . . xik with i1 < i2 < · · · < ik . ¤
Let ϕ be a quadratic form on V over F and v0 ∈ V a vector such that ϕ(v0 ) = 1. For
every v ∈ V , the element −vv0 in the even Clifford algebra C0 (ϕ) satisfies the quadratic
equation
(30.3) (−vv0 )2 + bϕ (v0 , v)(−vv0 ) + ϕ(v) = 0.
Choose a subspace U ⊂ V such that V = F v0 ⊕ U . Let J be the ideal of the tensor
algebra T (U ) generated by the elements v ⊗ v + bϕ (v0 , v)v + ϕ(v) for all v ∈ U .
Lemma 30.4. With U as above, the F -algebra homomorphism α : T (U )/J → C0 (ϕ)
defined by α(v + J) = −vv0 is an isomorphism.
Proof. By Lemma 30.2, we have dim T (U )/J ≤ 2dim U = dim C0 (ϕ). As α is surjec-
tive, it must be an isomorphism. ¤
Theorem 30.5. Let D be a finite dimensional division F -algebra and ϕ an irreducible
quadratic form over F . Then DF (ϕ) is not a division algebra if and only if there is an
F -algebra homomorphism C0 (ϕ) → D.
Proof. Scaling ϕ, we may assume that there is v0 ∈ V satisfying ϕ(v0 ) = 1 where
V = Vϕ . We use the decomposition V = F v0 ⊕ U as above and set
l(v) = bϕ (v0 , v) for every v ∈ U.
Claim: Suppose that DF (ϕ) is not a division algebra. Then there is an F -linear map
f : U → D satisfying the equality of quadratic maps
(30.6) f 2 + lf + ϕ = 0.
(We view the left hand side as the quadratic map v 7→ f (v)2 + l(v)f (v) + ϕ(v) on U ).
If we establish the claim then the map f extends to an F -algebra homomorphism T (U )/J →
D and, by Lemma 30.4, we get an F -algebra homomorphism C0 (ϕ) → D as needed.
We prove the claim by induction on dim U . Suppose that dim U = 1, i.e., U = F v
for some v. By Example 22.2, we have F (ϕ) ' C0 (ϕ) = F ⊕ F x with x satisfying the
quadratic equation x2 + ax + b = 0 with a = l(v) and b = ϕ(v) by equation (30.3).
Since DF (ϕ) is not a division algebra, there exists a nonzero element d0 + dx ∈ DF (ϕ) with
d, d0 ∈ D such that (d0 + dx)2 = 0 or equivalently d0 2 = bd2 and dd0 + d0 d = ad2 . Since D
is a division algebra, we have d 6= 0. Then the element d0 d−1 in D satisfies
(d0 d−1 )2 − a(d0 d−1 ) + b = 0.
Therefore, the assignment v 7→ −d0 d−1 gives rise to the desired map f : U → D.
Now consider the general case dim U ≥ 2. Choose a decomposition
U = F v1 ⊕ F v 2 ⊕ W
for some nonzero v1 , v2 ∈ U and a subspace W ⊂ U and set V 0 = F v0 ⊕ F v1 ⊕ W ,
U 0 = F v1 ⊕ W so that V 0 = F v0 ⊕ U 0 . Consider the quadratic form ϕ0 on the vector space
VF0 (t) over the function field F (t) defined by
ϕ0 (av0 + bv1 + w) = ϕ(av0 + bv1 + btv2 + w).
We show that the function fields F (ϕ) and F (t)(ϕ0 ) are isomorphic over F . Indeed,
consider the injective F -linear map θ : V ∗ → V 0 ∗F (t) taking a linear functional z to the
functional z 0 defined by z 0 (av0 + bv1 + w) = z(av0 + bv1 + btv2 + w). The map θ identifies
the ring S • (V ∗ ) with a graded subring of S • (V 0 ∗F (t) ) so that ϕ is identified with ϕ0 . Let x1
and x2 be the coordinate functions of v1 and v2 in V respectively and x01 the coordinate
function of v1 in V 0 . We have x1 = x01 and x2 = tx01 in S 1 (V 0 ∗F (t) ). Therefore, the
localization of the ring S • (V ∗ ) with respect to the multiplicative system F [x1 , x2 ] \ {0}
coincides with the localization of S • (V 0 ∗F (t) ) with respect to F (t)[x01 ] \ {0}. Note that
F [x1 , x2 ] ∩ (ϕ) = 0 and F (t)[x01 ] ∩ (ϕ0 ) = 0. It follows that the localizations S • (V ∗ )(ϕ) and
S • (V 0 ∗F (t) )(ϕ0 ) are equal. As the function fields F (ϕ) and F (t)(ϕ0 ) coincide with the degree
0 components of the quotient fields of their respective localizations, the assertion follows.
Let l0 (v) = b0ϕ (v0 , v), so
l0 (av0 + bv1 + w) = l(av0 + bv1 + btv2 + w).
Applying the induction hypothesis to the quadratic form ϕ0 over F (t) and the F (t)-algebra
DF (t) produces an F (t)-linear map f 0 : UF0 (t) → DF (t) satisfying
2
(30.7) f 0 + l0 f 0 + ϕ0 = 0.
Consider the F [t]-submodule X = f 0 (UF0 [t] ) in DF [t] . By Lemma 30.2, the F [t]-subalgebra
generated by X is a finitely generated F [t]-module. It follows from Lemma 30.1 that, after
applying an inner automorphism of DF (t) , we have f 0 (v) ∈ DF [t] for all v. Considering the
highest degree terms of f 0 (with respect to t) and taking into account the fact that D is a
division algebra, we see that deg f 0 ≤ 1, i.e., f 0 = g +ht for two linear maps g, h : U 0 → D.
Comparison of degree 2 terms of (30.7) yields
h(v)2 + bl(v2 )h(v) + b2 ϕ(v2 ) = 0
for all v = bv1 + w. In particular, h is zero on W , therefore h(v) = bh(v1 ). Thus (30.7)
reads
¡ ¢2 ¡ ¢
(30.8) g(v) + bth(v1 ) + l(bv1 + btv2 ) g(v) + bth(v1 ) + ϕ(v + btv2 ) = 0
for every v = bv1 + w. Let f : U → D be the F -linear map defined by the formula
f (bv1 + cv2 + w) = g(bv1 + w) + ch(v1 ).
Substituting c/b for t in (30.8), we see that (30.6) holds on all vectors bv1 + cv2 + w with
b 6= 0 and therefore holds as an equality of quadratic maps. The claim is proven.
We now prove the converse. Suppose that there is an F -algebra homomorphism
s : C0 (ϕ) → D.
¡ Consider ¢ the two F -linear maps p, q : V → D given by p(v) = s(vv0 )
and q(v) = s vv0 − l(v) . We have
¡ ¢ ¡ ¢
p(v)q(v) = s (vv0 )2 − l(v)vv0 = s ϕ(v) = ϕ(v)
by equation (30.3). It follows that p and q are injective maps if ϕ is anisotropic. The
maps p and q stay injective over any field extension. Let L/F be a field extension such
that ϕL is isotropic (e.g., L = F (ϕ)). Then if v 0 ∈ VL is a nonzero isotropic vector, we
have p(v 0 )q(v 0 ) = ϕ(v 0 ) = 0 but p(v 0 ) 6= 0 and q(v 0 ) 6= 0. It follows that DL is not a
division algebra.
It remains to consider the case when ϕ is isotropic. We first show that every isotropic
vector v ∈ V belongs to rad bϕ . Suppose this is not true. Then there is a u ∈ V satisfying
bϕ (v, u) 6= 0. Let H be the 2-dimensional subspace generated by v and u. The restriction
of ϕ on H is a hyperbolic plane. Let w ∈ V be a nonzero vector orthogonal to H and let
a = ϕ(w). Then
M2 (F ) = C(−aH) = C0 (F w ⊥ H) ⊂ C0 (ϕ)
by Proposition 11.4. The image of the matrix algebra M2 (F ) under s is isomorphic to
M2 (F ) and therefore contains zero divisors, a contradiction proving the assertion.
Let V 0 be a subspace of V satisfying V = rad ϕ ⊕ V 0 . As every isotropic vector belongs
to rad bϕ , the restriction ϕ0 of ϕ on V 0 is anisotropic. The natural map C0 (ϕ) → C0 (ϕ0 )
∼
induces an isomorphism C0 (ϕ)/J → C0 (ϕ0 ), where J = (rad ϕ)C1 (ϕ). Since x2 = 0 for
every x ∈ rad ϕ, we have s(J) = 0. Therefore, s induces an F -algebra homomorphism
s0 : C0 (ϕ0 ) → D. By the anisotropic case, D is not a division algebra over F (ϕ0 ). Since
F (ϕ) is a field extension of F (ϕ0 ), the algebra DF (ϕ) is also not a division algebra. ¤
Corollary 30.9. Let D be a division F -algebra of dimension less than 22n and ϕ a
non-degenerate quadratic form of dimension at least 2n + 1 over F . Then DF (ϕ) is also a
division algebra.
Proof. Let ψ be a non-degenerate subform of ϕ of dimension 2n+1. As F (ψ)(ϕ)/F (ψ)
is a purely transcendental extension by Proposition 22.9, we may replace ϕ by ψ and
assume that dim ϕ = 2n + 1. By Proposition 11.6, the algebra C0 (ϕ) is simple of dimen-
sion 22n . If DF (ϕ) is not a division algebra then there is an F -algebra homomorphism
C0 (ϕ) → D by Theorem 30.5. This homomorphism must be injective as C0 (ϕ) is simple.
But this is impossible by dimension count. ¤
Corollary 30.10. Let D be a division F -algebra and let ϕ be a non-degenerate qua-

dratic form over F satisfying:
(1) If dim ϕ is odd or ϕ ∈ Iq (F ) \ Iq2 (F ) then C0 (ϕ) is not a division algebra.
(2) If ϕ ∈ Iq2 (F ) then C + (ϕ) is not a division algebra over F (cf. Remark 13.9).
Then DF (ϕ) is a division algebra.
Proof. If DF (ϕ) is not a division algebra, there is an F -algebra homomorphism
f : C0 (ϕ) → D by Theorem 30.5. If ϕ ∈ Iq2 (F ) we have C0 (ϕ) ' C + (ϕ) × C + (ϕ) by
Remark 13.9. Thus in every case the image of f lies in a non-division subalgebra of D.
Therefore, D is not a division algebra, a contradiction. ¤
Corollary 30.11. Let D be a division F -algebra and let ϕ ∈ Iq3 (F ) be a nonzero
quadratic form. Then DF (ϕ) is a division algebra.
Proof. By Theorem 14.3, the Clifford algebra C(ϕ) is split. In particular, C + (ϕ) is
not a division algebra. The statement follows now from Corollary 30.10. ¤
CHAPTER V
Bilinear and Quadratic Forms and Algebraic Extensions
31. Structure of the Witt ring
In this section, we investigate the structure of the Witt ring of non-degenerate sym-
metric bilinear forms. For fields F whose level s(F ) is finite, i.e., non-formally real fields,
the ring structure is quite simple. The Witt ring of such a field has a unique prime
ideal, viz., the fundamental ideal and W (F ) (as an abelian group) has exponent 2s(F ).
As s(F ) = 2n for some non-negative integer this means that the Witt ring is 2-primary
torsion. The case of formally real fields F , i.e., fields of infinite level, is more involved.
Orderings on such a field give rise to prime ideals in W (F ). The torsion in W (F ) is still
2-primary, but this is not as easy to show. Therefore, we do the two cases separately.
31.A. Non-formally real fields. We consider the case of non-formally real fields
first.
A field F is called quadratically closed if F = F 2 . For example, algebraically closed
fields are quadratically closed. A field of characteristic two is quadratically closed if and
only if it is perfect. The quadratic closure of the rationals Q is the field of complex
constructible numbers. Over a quadratically closed field, the structure of the Witt ring is
very simple. Indeed, we have
Lemma 31.1. For a field F the following are equivalent:
(1) F is quadratically closed.
(2) W (F ) = Z/2Z.
(3) I(F ) = 0.
Proof. As W (F )/I(F ) = Z/2Z, we have W (F ) ' Z/2Z if and only if I(F ) = 0 if
and only if h1, −ai = 0 in W (F ) for all a ∈ F × if and only if a ∈ F ×2 for all a ∈ F × . ¤
Example 31.2. (1) Let F be a finite field with char F = p and |F | = q. If p = 2 then
F = F 2 and F is quadratically closed. So suppose that p > 2. Then F ×2 ' F × /{±1} so
|F × /F ×2 | = 2 and |F 2 | = 21 (q + 1). Let F × /F ×2 = {F ×2 , aF ×2 }. Since the finite sets
F2 and {a − y 2 | y ∈ F }
both have 12 (q + 1) elements, they intersect non-trivially. It follows that every element in
F is a sum of two squares. We have −1 ∈ F ×2 if and only if q ≡ 1 mod 4.
If q ≡ 3 mod 4 then −1 ∈ / F ×2 and the level s(F ) of F is 2. We may assume that
a = −1. Then h1, 1, 1i = h1, −1, −1i = h−1i in W (F ) so W (F ) equals {0, h1i, h−1i, h1, 1i}
and is isomorphic to the ring Z/4Z.
121
122 V. BILINEAR AND QUADRATIC FORMS AND ALGEBRAIC EXTENSIONS
If q ≡ 1 mod 4 then −1 is a square and W (F ) equals {0, h1i, hai, h1, ai} and is
isomorphic to the group ring Z/2Z[F × /F ×2 ].
(2) If F is not formally real with char F 6= 2 then s = s(F ) is finite so the symmetric
bilinear form (s + 1)h1i is isotropic hence universal by Corollary 1.26.
It follows by the above that any field F of positive characteristic level s(F ) = 1 or 2.
In general, if F is not formally real, s(F ) = 2n by Corollary 6.8. There exist fields of level
2m for each m ≥ 1.
Lemma 31.3. Let 2m ≤ n < 2m+1 . Suppose ¡
m
¢that mF satisfies s(F ) > 2 , e.g., F is
formally real, and ϕ = (n + 1)h1iq . Then s F (ϕ) = 2 .
Proof. As s(F
¡ ) > ¢ 1, the characteristic of F is not two. Since ϕF (ϕ) is isotropic,
m
it¡ follows
¢ that s F (ϕ) ≤ 2 by Corollary 6.8. If ϕ was isotropic over F then s(F ) =
m
s F (ϕ) ≤ 2 as F (ϕ)/F is purely transcendental
¡ ¢ by Proposition
¡ ¢22.9. This contradicts
the hypothesis. So ϕ is anisotropic. If s F (ϕ) < 2m then 2m h1i F (ϕ) is a Pfister form
as char F 6= 2 hence is hyperbolic. It follows that 2m = dim 2m h1i ≥ dim ϕ > 2m by the
Subform Theorem 22.5, a contradiction. ¤
The ring structure of W (F ) due to Pfister (cf. [113]) is given by the following:
Proposition 31.4. Let F be non-formally real with s(F ) = 2n . Then
¡ ¢
(1) Spec W (F ) = {I(F )}
(2) W (F
¡ ) is a¢ local ring
¡ of Krull
¢ dimension zero with maximal ideal I(F ).
(3) nil W (F ) = rad W (F ) = zd(F ) = I(F ).
(4) W (F )× = {b | dim b is odd}
(5) W (F ) is connected, i.e., 0 and 1 are the only idempotents in W (F ).
(6) W (F ) is a 2-primary torsion group of exponent 2s(F ).
(7) W (F ) is artinian if and only if it is noetherian if and only if |F × /F ×2 | is finite
if and only if W (F ) is a finite ring.
Proof. Let s = s(F ). The integer 2s is the smallest integer such that the bilinear
Pfister form 2sh1ib is metabolic hence zero in the Witt ring. Therefore, 2n+1 hai = 0 in
W (F ) for every a ∈ F × . It follows that W (F ) is 2-primary torsion of exponent 2n+1 , i.e.,
(6) holds. As
hhaiin+2 = hha, . . . , aii = hha, −1, . . . , −1ii = 2n+1 hhaii = 0
in W (F ) for every a ∈ F × by Example 4.16, we have I(F ) lies in every prime ideal. Since
W (F )/I(F ) ' Z/2Z, the fundamental ideal I(F ) is maximal hence is the only prime
ideal. This proves (1). As I(F ) is the only prime ideal (2) − (5) follows easily.
Finally, we show (7). Suppose that W (F ) is noetherian. Then I(F ) is a finitely
generated W (F )-module so I(F )/I 2 (F ) is a finitely generated W (F )/I(F )-module. As
F × /F ×2 ' I(F )/I 2 (F ) by Proposition 4.13 and Z/2Z ' W (F )/I(F ), we have F × /F ×2 is
finite. Conversely, suppose that F × /F ×2 is finite. By (2.6), we have a ring epimorphism
Z[F × /F ×2 ] → W (F ). As the group ring Z[F × /F ×2 ] is noetherian, W (F ) is noetherian.
As 2sW (F ) = 0 and W (F ) is generated by the classes of 1-dimensional forms, we see that
|W (F )| ≤ |F × /F ×2 |2s . Statement (7) now follows easily. ¤
We turn to formally real fields, i.e., those fields of infinite level. In particular, for-
mally real fields are of characteristic zero, so the theories of symmetric bilinear forms and
quadratic forms merge. The structure of the Witt ring of a formally real field is more
complicated as well as more interesting. We shall use the basic algebraic and topological
structure of formally real fields which can be found in Appendices §95 and §96. Recall
that a formally field F is called euclidean if every element in F × is a square or minus a
square. So F is euclidean if and only if F is formally real and F × /F ×2 = {F ×2 , −F ×2 }.
In particular, every real-closed field (cf. 95) is euclidean. Sylvester’s Law of Inertia for
real-closed fields generalizes to euclidean fields.
Proposition 31.5. (Sylvester’s Law of Inertia) Let F be a field. Then the following
are equivalent:
(1) F is euclidean.
(2) F is formally real and if b is a non-degenerate symmetric bilinear form, there
exists unique non-negative integers m, n such that b ' mh1i ⊥ nh−1i.
(3) W (F ) ' Z as rings.
(4) F 2 is an ordering of F .
Proof. (1) ⇒ (2): As F is formally real, char F = 0 so every bilinear form is
diagonalizable. Since F × /F ×2 = {F ×2 , −F ×2 }, every non-degenerate bilinear form is
isometric to mh1i ⊥ nh−1i for some non-negative integers n and m. The integers n and
m are unique by Witt Cancellation 1.29.
(2) ⇒ (3): By (2) every anisotropic quadratic form is isometric to rh1i for some unique
integer r. As F is formally real every such form is anisotropic.
(3) ⇒ (4): Let sgn : W (F ) → Z be the isomorphism. Then sgnh1i = 1 so h1i has
infinite order, hence F is formally real. Let a ∈ F . Then sgnhai = n for some integer
n. Thus hai = nh1i in W (F ). In particular n is odd. Taking determinants, we must
have aF ×2 = ±F ×2 . It follows that F × /F ×2 = {F ×2 , −F ×2 }. As F is formally real,
F 2 + F 2 ⊂ F 2 hence F 2 is an ordering.
(4) ⇒ (1): As F has an ordering, it is formally real. As F 2 is an ordering, F = F 2 ∪(−F 2 )
with −1 ∈ / F 2 , so F is euclidean. ¤
31.B. Pythagorean fields. We turn to a class of fields important in the study of
formally real fields that generalize quadratically closed fields.
Definition 31.6. Let F be a euclidean field. If b is a non-degenerate symmetric
bilinear form then b ' mh1i ⊥ nh−1i for unique non-negative integers n and m. The
integer m − n is called the signature of b and denoted sgn b. This induces an isomorphism
sgn : W (F ) → Z taking the Witt class of b to sgn b called the signature map.
Let
¡ ¢ [ ¡ ¢
D ∞h1i := D nh1i = {x | x is a nonzero sum of squares in F }
n
¡ ¢ ¡ ¢
e ∞h1i := D ∞h1i ∪ {0}.
D
A field F is called
¡ a ¢pythagorean field if every sum of squares of elements in F is itself a
e
square, i.e., D ∞h1i = F 2 and if char F = 2 then F is quadratically closed, i.e., perfect.
Remark 31.7. A field F of characteristic different from two is pythagorean if and

only if every sum of two squares F is a square.
Example 31.8. Let F be a field.
(1) Every euclidean field is pythagorean.
(2) Let F be a field of characteristic different from two and K = F ((t)), a Laurent series
field over F . Then K is the quotient field of F [[t]], a complete discrete valuation ring. If
F is formally real then so is K as nh1i is anisotropic over K for all n by Lemma 19.4.
Suppose that F is formally real and pythagorean. If xi ∈ K × for i = 1, 2 then there exist
integers mi such that xi = tmi (ai + tyi ) with ai ∈ F × and yi ∈ F [[t]] for i = 1, 2. Suppose
that m1 ≤ m2 then x21 + x22 = t2m1 (c + tz) with z ∈ F [[t]] and c = a21 if m1 < m2 and
c = a21 + a22 if m1 = m2 hence c is a square in F in either case. As K is formally real, c 6= 0
in either case. Hence c + tz is a square in K by Hensel’s Lemma. It follows that K is also
pythagorean. In particular, the finitely iterated Laurent series field Fn = F ((t1 )) · · · ((tn ))
as well as the infinitely iterated Laurent series field F∞ = colim Fn = F ((t1 )) · · · ((tn )) · · ·
are formally real and pythagorean if F is.
¡ ¢
(3) If F is not formally real and char F 6= 2 then F = D e ∞h1i by Example 31.2(2). It
follows that if F is not formally real then F is pythagorean if and only if it is quadratically
closed.
(4) Let K = F ((t)) with char F = 0 and F × /F ×2 = {ai F ×2 | i ∈ I}. It follows by
Hensel’s Lemma that
K × /K ×2 = {ai K ×2 | i ∈ I} ∪ {ai tK ×2 | i ∈ I}
and from Lemma 19.4 that this is a disjoint union and ai K ×2 = aj K ×2 if and only if i = j.
In particular, if F is not formally real then Laurent series field K is not pythagorean as
t is not a square.
Exercise 31.9. Let F be a formally real pythagorean field and let b be a bilinear
form over F . Prove that the set D(b) is closed under addition.
Proposition 31.10. Let F be a field. Then the following are equivalent:
(1) F is pythagorean.
(2) I(F ) is torsion-free.
(3) There are no anisotropic torsion binary bilinear forms over F .
Proof. (1) ⇒ (2): If s(F ) is finite then F is quadratically closed so W (F ) = {0, h1i}
and I(F ) = 0. Therefore, we may assume that F is formally real. We show in this case
that W (F ) is torsion-free. Let b be an anisotropic bilinear form over F that is torsion
in W (F ), say mb = 0 in W (F ) for some positive integer m. As b is diagonalizable by
Corollary 1.20, suppose that b ' ha . . . , an i with ai ∈ F × . The form mbi is isotropic so
P1 , P
there exists a nontrivial equation j i ai x2ij = 0 in F . As F is pythagorean, there exist
P
xi ∈ F satisfying x2i = j x2ij . Since F is formally real not all the xi can be zero. Thus
(x1 , . . . , xn ) is an isotropic vector for b, a contradiction.
¡ ¢
(3) ⇒ (1): Let 0 6= z ∈ D 2h1i . Then 2hhzii = 0 in W (F ) by Corollary 6.6. By
assumption, hhzii = 0 in W (F ) hence z ∈ F ×2 . ¤
Corollary 31.11. A field F is formally real and pythagorean if and only if W (F ) is

torsion-free.
Proof. Suppose that W (F ) is torsion-free. Then I(F ) is torsion-free so F is pythagorean.
As h1i is not torsion, s(F ) is infinite hence F is formally real.
Conversely, suppose that F is formally real and pythagorean. Then the proof of
(1) ⇒ (2) in Proposition 31.10 shows that W (F ) is torsion-free. ¤
Lemma 31.12. The intersection of pythagorean fields is pythagorean.
T
Proof. Let F = I Fi with each Fi pythagorean. If z = x2 + y 2 with x, y ∈ F then
2
for each
T i ∈ I there exist zi ∈ Fi with zi =2 z. In particular, zi = ±zj for all i, j ∈ I. Thus
zj ∈ I Fi = F for every j ∈ I and z = zj . ¤
Exercise 31.13. Let K/F be a finite extension. Show √ if K is pythagorean√so is F .
(Hint: If char F 6= 2 and a = 1 + x2 ∈ F \ F 2 , let z = a + a ∈ K. Show z ∈ F ( a)2 but
/ F 2 .)
NF (√a)/F (z) ∈
Let F be a field and K/F an algebraic extension. We call K a pythagorean closure of F
if K is pythagorean and if F ⊂ E $ K is an intermediate field then E is not pythagorean.
If Fe is an algebraic closure of F then the intersection of all pythagorean fields between
F and Fe is pythagorean by the lemma. Clearly, this is a pythagorean closure of F . In
particular, a pythagorean closure is unique (after fixing an algebraic closure). We shall
denote the pythagorean closure of F by Fpy . If F is not a formally real field then Fpy is
just the quadratic closure of F , i.e., a quadratically closed field K algebraic over F such
that if F ⊂ E $ K then E is not quadratically closed.
Exercise 31.14. Let E be a pythagorean closure of a field F . Prove that E/F is an
excellent extension. (Hint: in the formally real case use Exercise 31.9 to show that for
any quadratic form ϕ over F the form (ϕE )an over E takes values in F .)
We next show how to construct the pythagorean closure of a field.

Definition 31.15. Let F be a field with Fe an algebraic closure. If K/F is a finite
extension in Fe then we say K/F is admissible if there exists a tower
F = F0 ⊂ F1 ⊂ · · · ⊂ Fn = K where
(31.16) √
Fi = Fi−1 ( zi−1 ) with zi−1 ∈ D(2h1iFi−1 )
from F to K.
Remark 31.17. If F is a formally real field and K an admissible extension of F then
K is formally real by Theorem 95.3 in §95.
Lemma 31.18. Suppose that char F 6= 2. Let L be the union of all admissible extensions
over F . Then L = Fpy . If F is formally real so is Fpy .
Proof. Let Fe be a fixed algebraic closure of F . If E and K are admissible extensions
of F then the compositum of EK of E and K is also an admissible extension. It follows
that L is a field. If z ∈ L satisfies z = x2 + y 2 , x, y ∈ L, then there exist admissible
√
extensions E and K of√ F with x√∈ E and y ∈ K. Then EK( z) is an admissible
extension of F hence z ∈ EK( z) ⊂ L. Therefore, L is pythagorean. Let M be
pythagorean with F ⊂ M ⊂ Fe. We show L ⊂ M . Let K/F be admissible. Let (31.16)
be a tower from F to K. By induction, we may assume that Fi ⊂ M . Therefore, zi ∈ M 2
hence Fi+1 ⊂ M . Consequently, K ⊂ M . It follows that L ⊂ M so L = Fpy . If F is
formally real then so is L by Remark 31.17. ¤
If F is an arbitrary field then the quadratic closure of F can also be constructed by
taking the union of all square root towers
√ ×
F = F0 ⊂ F1 ⊂ · · · ⊂ Fn = K where Fi = Fi−1 ( zi−1 ) with zi−1 ∈ Fi−1
over F .
Notation 31.19. Let
Wt (F ) := {b ∈ W (F ) | there exists a positive integer n such that nb = 0},
the additive torsion in W (F ). It is an ideal in W (F ).
¡ ¢
Recall if K/F is a field extension then W (K/F ) := ker rK/F : W (F ) → W (K) .
¡ ¢ √
Lemma 31.20. Let z ∈ D 2h1i \ F ×2 . If K = F ( z) then
¡ ¢
W (K/F ) ⊂ annW (F ) 2h1i .
Proof. It follows from the hypothesis that hhzii is anisotropic hence K/F is a qua-
dratic extension. As z is a sum of squares and not a square, char F 6= 2. Therefore, by
Corollary 23.7, we have W (K/F ) = hhziiW (F ). By Corollary 6.6, we have 2hhzii = 0 in
W (F ) and the result follows. ¤
31.C. Formally real fields. We have
Theorem 31.21. Let F be a formally real field.
(1) Wt (F ) is 2-primary, i.e., all torsion elements of W (F ) have exponent a power of
2.
(2) Wt (F ) = W (Fpy /F ).
Proof. As W (Fpy ) is torsion-free by Corollary 31.11, the torsion subgroup Wt (F ) lies
in W (Fpy /F ), so it suffices to show W (Fpy /F ) is a 2-primary torsion group. Let K be an
admissible extension of F as in (31.16). Since Fpy is the union of admissible extensions
by Lemma 31.18, it suffices to show W (K/F ) ¡is 2-primary
¢ torsion. By Lemma 31.20 and
n
induction, it follows that W (K/F ) ⊂ annW (F ) 2 h1i as needed. ¤
Lemma 31.22. Let F be a formally real field and b ∈ W (F ) satisfy 2n b 6= 0 in W (F )
for any n ≥ 0. Let K/F be an algebraic extension that is maximal with respect to bK not
having order a power of 2 in W (K). Then K is euclidean. In particular, sgn bK 6= 0.
Proof. Suppose K is not euclidean. As 2n h1i 6= 0, the field K is formally real. Since
euclidean, there√exists an x ∈ K × such that x ∈
K is not √ / (K × )2 ∪ −(K × )2 . In particular,
both K( x)/K and K( −x)/K are quadratic extensions. By choice of K, there exists
a positive integer n such that c := 2n bK satisfies cK(√x) and cK(√−x) are metabolic, hence
hyperbolic as char F 6= 2. By Corollary 23.7, there exist forms c1 and c2 over K satisfying
c ' hhxii ⊗ c1 ' hh−xii ⊗ c2 . As −xhhxii ' hhxii and xhh−xii ' hh−xii, we conclude
that xc ' c ' −xc and hence that 2c ' c ⊥ c ' xc ⊥ −xc. Thus 2c = 0 in W (K). This
means that bK is torsion of order 2n+1 , a contradiction. ¤
Proposition 31.23. The following are equivalent:
(1) F can be ordered, i.e., X(F ), the space of orderings of F , is not empty.
(2) F is formally real.
(3) Wt (F ) 6= W (F ).
(4) W (F ) is not a 2-primary torsion group.
(5) There exists an ideal A ⊂ W (F ) such that W (F )/A ' Z.
(6) There exists a prime ideal P in W (F ) such that char(W (F )/P) 6= 2.
¡ ¢
Moreover, if F is formally real then for any prime ideal P in W (F ) with char W (F )/P 6=
2, the set
PP := {x ∈ F × | hhxii ∈ P} ∪ {0}
is an ordering of F .
Proof. (1) ⇒ (2) is clear.
¡ ¢
(2) ⇒ (3): By assumption, −1 ∈ / DF nh1i for any n > 0 so h1i ∈ / Wt (F ).
(4) ⇒ (5): By assumption there exists b ∈ W (F ) not having order a power of 2. By
Lemma 31.22, there exists K/F with K euclidean. In particular, rK/F is surjective.
Therefore, A = W (K/F ) works by Lemma 31.22 and Sylvester’s Law of Inertia 31.5.
(6) ⇒ (1): By Proposition 31.4, the field F is formally real. We show that (6) implies
the last statement. This will also prove (1). Let P in W (F ) be a prime ideal satisfying
char(W (F )/P) 6= 2.
We must show
(i) PP ∪ (−PP ) = F .
(ii) PP + PP ⊂ PP .
(iii) PP · PP ⊂ PP .
(iv) PP ∩ (−PP ) = {0}.
(v) −1 6∈ PP .
Suppose that x 6= 0 and both ±x ∈ PP . Then hh−1ii = hh−xii + hhxii lies in P so
2h1i+P = 0 in W (F )/P, a contradiction. This shows (iv) and (v) hold. As hhx, −xii = 0
in W (F ), either hhxii or hh−xii lies in P, so (i) holds. Next let x, y ∈ PP . Then
hhxyii = hhxii + xhhyii lies in P so xy ∈ P which is (iii). Finally, we show that (ii) holds,
i.e., x + y ∈ PP . We may assume neither x nor y is zero. This implies that z := x + y 6= 0
else we have the equation hh−1ii = h1, x, −x, 1i = h1, −x, −y, 1i = hhxii + hhyii in W (F )
which implies that hh−1ii lies in P contradicting (v). Since h−x, −yi ' −zhh−xyii by
Corollary 6.6, we have
2h−zi = 2h−x, −y, zxyi = h−x, −y, zxy, −z, −zxy, zxyi = hhxii + hhyii − 2h1i − zhhxyii
in W (F ). As x, y ∈ PP and xy ∈ PP by (iii), it follows that 2hhzii ∈ P as needed. ¤
The proposition gives another proof of the Artin-Schreier Theorem that every formally
real field can be ordered.
In [95], Lewis gave an elegant proof for the structure of the Witt ring, part of which
we present in the following exercise.
Exercise 31.24. Let F be an arbitrary field. Show all of the following:
(1) (Leung) Define a polynomial pn (t) in F [t] as follows:
(
t(t2 − 22 )(t2 − 42 ) · · · (t2 − n2 ) if n is even
pn (t) =
(t2 − 12 )(t2 − 32 )(t2 − 52 ) · · · (t2 − n2 ) if n is odd.
If b is an n-dimensional non-degenerate symmetric bilinear form over F then
pn (b) = 0 in W (F ).
(2) W (F ) is integral over Z if F is formally real and Z/2r Z for some r if F is not
formally real. In particular, the Krull dimension of W (F ) is one if F is formally
real and zero if not.
(3) Using Exercise 33.14 below show that Wt (F ) is 2-primary.
¡ ¢
(4) W (F ) has no odd-dimensional zero divisors and if Wt (F ) 6= 0 then zd W (F ) =
I(F ). Moreover, W (F ) contains no nontrivial idempotents.
¡ ¢
(5) If F is formally real then Wt (F ) = nil W (F ) .
31.D. The Local-Global Principle. The main result of this subsection is Theorem
31.25 below due to Pfister (cf. [113]).
Let F be a formally real field and X(F ) the space of orderings. Let P ∈ X(F ) and
FP be the real closure of F at P (within a fixed algebraic closure). By Sylvester’s Law of
Inertia 31.5, the signature map defines an isomorphism sgn : W (FP ) → Z. In particular,
we have a signature map sgnP : W (F ) → Z given by sgnP = sgn ◦ rFP /F . This is a ring
homomorphism satisfying Wt (F ) ⊂ ker(rFP /F ) = ker(sgnP ). We let
¡ ¢
PP = ker(sgnP ) in Spec W (F ) .
Note if F ⊂ K ⊂ FP and b is a non-degenerate symmetric bilinear form then sgnP b =
sgnFP2 ∩K bK . In particular, if K is euclidean then sgnP b = sgn bK .
Theorem 31.25. (Local-Global Principle) The sequence
P
(rF /F )
Y
0 → Wt (F ) → W (F ) −→ W (FP )
X(F )
is exact.
Proof. We may assume that F is formally real by Proposition 31.4. We saw above
that Wt (F ) ⊂ ker(sgnP ) for every ordering P ∈ X(F ), so the sequence is a zero sequence.
Suppose that b ∈ W (F ) is not torsion of 2-power order. By Lemma 31.22, there exists a
euclidean field K/F with bK not of 2-power order. As K 2 ∈ X(K), we have P = K 2 ∩F ∈
X(F ). Thus sgnP b = sgn bK 6= 0. The result follows. ¤
Corollary 31.26. The map

¡ ¢
X(F ) −→ {P ∈ Spec W (F ) | W (F )/P ' Z} given by P 7→ PP
is a bijection.
Proof. Let P ⊂ W (F ) be a prime ideal such that W (F )/P ' Z. As in Proposition
31.23, let PP := {x ∈ F × | hhxii ∈ P} ∪ {0} ∈ X(F ).
Claim: P 7→ PP is the inverse, i.e., P = PPP and P = PPP .
If P ∈ X(F ) then certainly P ⊂ PPP , so we must have P = PPP as both are orderings.
→ Z maps hxi to
∼
By definition, we see that the composition W (F ) → W (F )/P −
sgnPP hxi. Hence ker(sgnPP ) = P. ¤
The spectrum of the Witt ring can now be determined. This result is due to Lorenz-
Leicht (cf. [96]) and Harrison (cf. [51]).
¡ ¢
Theorem 31.27. Spec W (F ) consists of
(1) The fundamental ideal I(F ).
(2) PP with P ∈ X(F ).
(3) PP,p := PP + pW (F ) = sgn−1 P (pZ), p an odd prime, with P ∈ X(F ).
Moreover, all these ideals are different. The prime ideals in (1) and (3) are the maximal
ideals of W (F ). If F is formally real then the ideals in (2) are the minimal primes of
W (F ) and PP ⊂ PP,p ∩ I(F ) for all P ∈ X(F ) and for all odd primes p.
Proof. We may assume that F is formally real by Proposition 31.4. Let P be a prime
ideal in W (F ). Let a ∈ F × . As hha, −aii = 0 in W (F ) either hhaii ∈ P or hh−aii ∈ P.
In particular, hai ≡ ±h1i mod P. Hence W (F )/P is cyclic generated by h1i + P, so
W (F )/P ' Z or Z/pZ for p a prime. If x, y ∈ F × then hxi and hyi are units in W (F ),
so do not lie in P. Suppose that W (F )/P ' Z/2Z. Then we must have hx, yi ∈ P for
all x, y ∈ F × hence P = I(F ). So suppose that W (F )/P 6' Z/2Z. By Proposition 31.23,
the set P = PP lies in X(F ). Since W (F )/PP ' Z, we have PP ⊂ P. Hence P = PP or
P = PP,p for a suitable odd prime. As each P ∈ X(F ) determines a unique PP and PP,p
by Corollary 31.23, the result follows. ¤
Corollary
¡ ¢31.28. If F is formally real then dim W (F ) = 1 and the map X(F ) →
Min Spec W (F ) given by P 7→ ker(sgnP ) is a homeomorphism.
Proof. As hh1ii does not lie in any minimal prime, for each a ∈ F × either a ∈ PP or
−a ∈ PP but not both where P ∈ X(F ). The sets H(a) := {P | −a ∈ P } form a subbase
for the topology of X(F ) (cf. §96). As a ∈ P for P ∈ X(F ) if and
¡ only if hhaii ∈ PP if
and only if PP lies in the basic open set {P | a ∈
/ P for P ∈ Min Spec W (F )}, the result
follows. ¤
The structure of the Witt ring in the formally real case due to Pfister (cf. [113]) can
now be shown.
Proposition 31.29. Let F be formally real. Then
¡ ¢ ¡ ¢
(1) nil W (F ) = rad W (F ) = Wt (F ).
(2) W (F )× = {b | sgnP b = ±1 for all P ∈ X(F )}

= {hai + c | a ∈ F × and¡ c ∈ I 2 (F
¢ ) ∩ Wt (F )}.
(3) If F is not pythagorean then¡ zd W¢(F ) = I(F ).
S
(4) If F is pythagorean then zd W (F ) = X(F ) PP $ I(F ).
(5) W (F ) is connected, i.e., 0 and 1 are the only idempotents in W (F ).
(6) W (F ) is noetherian if and only if F × /F ×2 is finite.
¡ ¢ ¡ ¢
Proof. (1): If P ∈ X(F ) then PP = ∩p PP,p so nil W (F ) = rad W (F ) . By the
Local-Global Principle 31.25, we have
Y \ \ \ ¡ ¢
Wt (F ) = ker( rFP /F ) = ker(sgnP ) = PP = PP,p = nil W (F ) .
P ∈X(F ) X(F ) X(F ) X(F )
¡ ¢
(2): We have sgnP W (F )× ⊂ {±1} for all P ∈ X(F ). Let b be a non-degenerate
symmetric bilinear form satisfying sgnP b = ±1 for all P ∈ X(F ). Choose a ∈ F such
that c := b − hai lies in I 2 (F ) using Proposition 4.13. In particular, sgnP b ≡ sgnP hai
mod 4 hence sgnP b = sgnP hai for all P ∈ X(F ). Consequently, sgnP c = 0 for all
P ∈ X(F ) so c is torsion by the Local-Global Principle 31.25. By (1), the form c is
nilpotent hence b ∈ W (F )× .
(3), (4): As the set of zero divisors is a saturated multiplicative set, it follows by commu-
tative algebra that it is a union of prime ideals.
Suppose that F is not pythagorean. Then Wt (F ) 6= 0 by Corollary 31.11. In particular,
n
2 b = 0 ∈ W (F ) for some b 6= 0 in W (F ) and n ≥ 1 by Theorem 31.21. Thus hh−1ii
is ¡a zero ¢divisor. As I(F ) is the only prime ideal containing hh−1ii, we have I(F ) ⊂
zd W (F ) . Since ¡nh1i is¢not a zero divisor for ¡ any odd ¢ integer n by Theorem 31.21, no
PP,p can lie in zd W (F ) . It follows that zd W (F )) = I(F ), since PP ⊂ I(F ) for all
P ∈ X(F ).
Suppose that F is pythagorean. Then W (F ) is torsion-free so nh1i is¡ not a¢ zero-
divisor for any nonzero integer n. In particular, no maximal ideal lies in zd W (F ) . Let
P ∈ X(F ) and b ∈ PP . Then b is diagonalizable so we have b ' ha1 , . . . , an , b1 , . . . bn i
with ai , −bj ∈ P for all i, j. Let c = hha1 b1 , . . . , an bn ii. Then c is non-zero in W¡ (F ) as
¢
sgnP c = 2n . As hh−ai bi ii¡ · c = 0¢in W (F ) for all i, we have b ·¡c = 0 hence
¢ b ∈ zd W (F ) .
Consequently, PP ⊂ zd W (F ) for all P ∈ X(F ) hence zd W (F ) is the union of the
minimal primes.
(5): If the result is false then ¡ 1 =¢ e1 + e2 for some nontrivial idempotents e1 , e2 . As
e1 e2 = 0, we have e1 , e2 ∈ zd W (F ) ⊂ I(F ) which implies 1 ∈ I(F ), a contradiction.
(6): This follows by the same proof for the analogous result in Proposition 31.4. ¤
Let It (F ) := Wt (F ) ∩ I(F ).
¡ Proposition
¢ 31.30. If F is formally real then Wt (F ) is generated by hhxii with x ∈
D ∞h1i , i.e., Wt (F ) = It (F ) is generated by torsion 1-fold Pfister forms.
¡ n ¢
Proof. Let b ∈ Wt (F ). Then 2n b = 0 for some integer n
¡ n ¢> 0. Thus b ∈ annW (F ) 2 h1i .
By Corollary 6.23, there exist binary forms di ∈ annW (F ) 2 h1i satisfying b = d1 +· · ·+dm
in W (F ). The result follows. ¤
31.E. The Witt ring up to isomorphism. Because I(F ) is the unique ideal of
index two in W (F ), we can deduce the following due to Cordes and Harrison (cf. [28]):
Theorem 31.31. Let F and K be two fields. Then W (F ) and W (K) are isomorphic
as rings if and only if W (F )/I 3 (F ) and W (K)/I 3 (K) are isomorphic as rings.
Proof. The fundamental ideal is the unique ideal of index two in its Witt ring by
Theorem 31.27. Therefore any ring isomorphism W (F ) → W (K) induces a ring isomor-
phism W (F )/I 3 (F ) → W (K)/I 3 (K).
Conversely, let g : W (F )/I 3 (F ) → W (K)/I 3 (K) be a ring isomorphism. As the
fundamental ideal is the unique ideal of index two in its Witt ring by Theorem 31.27, the
map g induces an isomorphism I(F )/I 3 (F ) → I(K)/I 3 (K). From this and Proposition
4.13, it follows easily that g induces an isomorphism h : F × /F ×2 → K × /K ×2 .
We adopt the following notation. For a coset α = xK ×2 , write hαi and hhαii for the
forms hxi and hhxii in W (K) respectively. We also write s(a) for h(aF ×2 ). Note that
s(ab) = s(a)s(b) for all a, b ∈ F × .
By construction,
¡ ¢
g hhaii + I 3 (F ) ≡ hhs(a)ii mod I 2 (K)/I 3 (K).
As g(1) = 1, plugging in a = −1, we get hs(−1)i = h−1i. In particular,
(31.32) hs(1)i + hs(−1)i = h1i + h−1i = 0 ∈ W (K).
Since g is a ring homomorphism, we have

¡ ¢ ¡ ¢ ¡ ¢
g hha, bii + I 3 (F ) = g hhaii + I 3 (F ) · g hhbii + I 3 (F )
= hhs(a)ii · hhs(b)ii + I 3 (K)
= hhs(a), s(b)ii + I 3 (K).
for every a, b ∈ F × .
If a + b 6= 0, we have hha, bii ' hha + b, ab(a + b)ii by Lemma 4.15(3). Therefore,
¡ ¢
hhs(a), s(b)ii ≡ hhs(a + b), s ab(a + b) ii mod I 3 (K).
By Theorem 6.20, these two 2-fold Pfister forms are equal in W (K). Therefore,
¡ ¢
(31.33) hs(a)i + hs(b)i = hs(a + b)i + hs ab(a + b) i
in W (K).
Let F be the free abelian group with basis the set of isomorphism classes of 1-
dimensional forms hai over F . It follows from Theorem 4.8 and equations (31.32) and
(31.33) that the map F → W (K) taking hai to hs(a)i gives rise to a homomorphism
s : W (F ) → W (K). Interchanging the roles of F and K, we have, in a similar fashion, a
homomorphism W (K) → W (F ) which is the inverse of s. ¤
32. Addendum on torsion

¡ ¢
We know by Corollary 6.26 that if b ∈ annW (F ) 2h1i , i.e., if 2b = 0 in W (F ) that
b ' d1 ⊥ · · · ⊥ dn where each bi is a binary form annihilated by 2. In particular, if b is
an anisotropic bilinear Pfister
¡ form such that 2b = 0 in W (F ) then the pure subform b0
¢
of b satisfies D(b0 ) ∩ D 2h−1i 6= ∅. In general, if 2n b = 0 in W (F ) with n > 1, then b
is not isometric to binary forms annihilated by 2n nor does the pure subform of a torsion
bilinear Pfister form represent a totally negative element. In this addendum, we construct
a counterexample due to Arason and Pfister (cf. [4]). We use the following variant of the
Cassels-Pfister Theorem 17.3.
Lemma 32.1. Let char F 6= 2. Let ϕ = ha1 , . . . , an iq be anisotropic over F (t) with
a1 , . . . , an ∈ F [t] all satisfying deg ai ≤ 1. Suppose that 0 6= q ∈ D(ϕF (t) ) ∩ F [t]. Then
there exist polynomials f1 , . . . , fn ∈ F [t] such that q = ϕ(f1 , . . . , fn ), i.e., F [t] ⊗F ϕ
represents q.
Proof. Let ψ ' h−qi ⊥ ϕ and let
Q := {f = (f0 , . . . , fn ) ∈ F [t]n+1 | bψ (f, f ) = 0}.
Choose f ∈ Q such that deg f0 is minimal. Assume that the result is false. Then
deg f0 > 0. Write fi = f0 gi + ri with ri = 0 or deg ri < deg f0 for each i using the
Euclidean Algorithm. So deg ri2 ≤ 2 deg f0 − 2 for all i. Let g = (1, g1 , . . . , gn ) and define
h ∈ F [t]n+1 by h = cf + dg with c = bψ (g, g) and d = −2bψ (f, g). We have
bψ (cf + dg, cf + dg) = c2 bψ (f, f ) + 2cdbψ (f, g) + d2 bψ (g, g) = 0
so h ∈ Q. Therefore,
h0 = bψ (g, g)f0 − 2bψ (f, g)g0 = bψ (f0 g − 2f, g) = −bψ (f + r, g),
so
n
X
f0 h0 = −f0 bψ (f + r, g) = −bψ (f + r, f − r) = bψ (r, r) = ai ri
i=1
which is not zero as ϕ is anisotropic. Consequently,
deg h0 + deg f0 ≤ max{deg ai } + 2 deg f0 − 2 ≤ 2 deg f0 − 1
i
as deg ai ≤ 1 for all i. This is a contradiction. ¤

¡ ¢
Lemma 32.2. Let F be a formally real field and x, y ∈ D ∞h1i . Let b = hh−t, x + tyii,
a 2-fold Pfister form over F (t). If b ' b1¡ ⊥ b2 ¢over F (t) with b1 and
¡ b2 binary
¢ torsion
forms over F (t) then there exists a z ∈ D ∞h1i such that x, y ∈ D hh−zii .
Proof. If x (respectively, y or xy) is a square, let z = y (respectively, z = x) to
finish. So we may assume¡ they¢ are not squares. As b is round, we may also¡ assume¢ that
0
d1 ' hhwii with w ∈ D ∞h1i by Corollary 6.6. In particular, D(bF ) ∩ D ∞h−1i 6= ∅
by Lemma 6.11. Thus, there exists a positive integer n such that b0 ⊥ nh1i is isotropic.
Let c = ht, −(x + yt)i ⊥ nh1i. We have t(x + yt) ∈ D(c). The form h1, −yi is anisotropic
as is nh1i,
P since F is formally real. If c is isotropic, then we would have an equation
2
−tf = gi − (x + yt)h2 in F [t] for some f, gi , h ∈ F [t]. Comparing leading terms
2
32. ADDENDUM ON TORSION 133
implies that y is a square. So c is anisotropic. By Lemma 32.1, there exist c, d, fi ∈ F [t]

satisfying
f12 + · · · + fn2 + tc2 − (x + yt)d2 = t(x + yt).
Since h1, −yi and nh1i are anisotropic and t2 occurs on the right hand side, we must have
c, d are constants and deg fi ≤ 1 for all i. Write fi = ai + bi t with ai , bi ∈ F for 1 ≤ i ≤ n.
Then n n n
X X X
2 2 2 2
ai = xd , 2 ai bi = −c + x + yd , and b2i = y.
i=1 i=1 i=1
2
If d = 0 then ai = 0 for all i and x = c is a square which was excluded. So d 6= 0. Let
n
X n
X Xn
z=4 a2i · b2i − 4( ai bi )2 = 4xyd2 − (x − c2 + yd2 )2 .
i=1 i=1 i=1
Applying the Cauchy-Schwarz Inequality
¡ in
¢ each real closure of F , we see that z is non-
negative in every ordering e
so z ∈ D ∞h1i . As xy is not a square, z 6= 0. As d 6= 0, we
¡ ¢
have xy ∈ D hh−zii . Expanding, one checks
z = 4xyd2 − (x − c2 + yd2 )2 = 4xc2 − (x − yd2 + c2 )2 .
¡ ¢ ¡ ¢
Thus x ∈ D hh−zii . As hh−zii is round, y ∈ D hh−zii also. ¤
Lemma 32.3. Let F0 be a formally real field and u, y ∈ D(∞h1iF¡0 ). Let ¢x = u + t2
in¡F = F¢0 (t). If there exists a z ∈ D(∞h1iF ) such that x, y ∈ D hh−zii then y ∈
D hh−uii .
Proof. We may assume that y is not a square. By assumption, we may write
z = (u + t2 )f12 − g12 = yf22 − g22 for some f1 , f2 , g1 , g2 ∈ F0 (t).
Multiplying this equation by an appropriate square in F0 (t), we may assume that z ∈ F [t]
and that f1 , g¡1 , f2 , g¢2 ∈ F0 [t] have no common nontrivial factor. As z is totally positive,
i.e., lies in D ∞h1i , its leading term must be totally positive in F0 . Consequently,
deg g1 ≤ 1 + deg f1 and deg g2 ≤ deg f2 .
1 1
It follows that deg z ≤ 1 + deg f1 . We have
2 2
deg z = deg f2 otherwise y ∈ F 2 , a
contradiction. Thus, we have
deg f2 ≤ 1 + deg f1 and deg(g1 ± g2 ) ≤ 1 + deg f1 .
¡ ¢
If deg (u + t )f1 − yf22 < 2 deg f1 + 2 then y would be a square in F0 , a contradiction.
2 2
So ¡ ¢
deg (u + t2 )f12 − yf22 = 2 + 2 deg f1 .
As (u + t2 )f12 − yf22 = g12 − g22 , we have deg(g1 ± g2 ) = 1 + deg f1 . It follows that either
f1 or g1 − g2 has a prime factor p of odd degree. Let F = F0 [t]/(p) and : F0 [t] → F be
the canonical map. Suppose that f 1 = 0. Then z = −g 21 in F . As z is a sum of squares
in F0 [t] (possibly zero), we must also have z is a sum of squares in F . But [F : F0 ] is odd
hence F0 is still formally real by Theorem 95.3 or Springer’s Theorem 18.5. Consequently,
2
we must have z = g 1 = 0. This implies that yf 2 = g 22 . As y cannot be a square in the
odd degree extension F of F0 by Springer’s Theorem 18.5, we must have f 2 = 0 = g 2 .
But there exist no prime p dividing f1 , f2 , g1 , and g2 . Thus p6 | f1 in F0 [t]. It follows that
2 2 2
g 1 = g 2 which in turn implies that (u + t )f 1 − yf 2 = 0. As f 1 6= 0, we have f 2 6= 0, so
we conclude that hu, 1, −yiF is isotropic.¡ As [F ¢: F0 ] is odd, hu, 1, −yi is isotropic over F0
by Springer’s Theorem 18.5, i.e., y ∈ D hh−uii as needed. ¤
We now construct the counterexample.
Example 32.4. We apply the above lemmas in the following case. Let F0 = Q(t1 )
and u = 1 and y = 3. The element y is a sum of three but not two squares in F0 by the
Substitution Principle 17.7. Let K = F0 (t2 ) and b = hh−t2 , 1+t21 +3t2 ii over K.
¡ Then the¢
Pfister form 4b is isotropic hence metabolic so 4b = 0 in W (K). As 1, 3t22 ∈ D hh−3t22 iiK
and 3 ∈/ D(2h1iQ(t1 ) ), the lemmas imply that b is not isometric to an orthogonal sum of
binary torsion forms. In particular, it also follows that the form b has the property
D(b0 ) ∩ D(∞h−1iK ) = ∅.
33. The total signature
We saw when F is a formally real field, the torsion in the Witt ring W (F ) is determined
by the signatures at the orderings on F . In this section, we view the relationship between
bilinear forms over a formally real field F and the totality of continuous functions on the
topological space X(F ) of orderings on F with integer values.
We shall use results in Appendices §95 and §96. Let F be a formally real field. The
space of orderings X(F ) is a boolean space, i.e., a totally disconnected compact Hausdorff
space with a subbase the collection of sets
(33.1) H(a) = HF (a) := {P ∈ X(F ) | −a ∈ P }.
Let b be a non-degenerate symmetric bilinear form over F . Then we define the total
signature of b to be the map
(33.2) sgn b : X(F ) → Z given by sgn b(P ) = sgnP b.
Theorem 33.3. Let F be formally real. Then

sgn b : X(F ) → Z
is continuous with respect to the discrete topology on Z. The topology on X(F ) is the
coarsest topology such that sgn b is continuous for all b.
Proof. As Z is a topological group, addition of continuous functions is continuous.
As any non-degenerate symmetric bilinear form is diagonalizable over a formally real field,
we need only prove the result for b = hai, a ∈ F × . But

¡ ¢−1  ∅ if n 6= ±1
sgnhai (n) = H(a) if n = −1

H(−a) if n = 1.
The result follows easily as the H(a) form a subbase. ¤
33. THE TOTAL SIGNATURE 135
¡ ¢
Let C X(F ), Z be the ring of continuous functions f : X(F ) → Z where Z has the
discrete topology. By the theorem, we have a map
¡ ¢
(33.4) sgn : W (F ) → C X(F ), Z given by b 7→ sgn b
called the total signature map. It is a ring homomorphism. The Local-Global Theorem
31.25 in this terminology states
Wt (F ) = ker(sgn).
¡ ¢
We turn to the cokernel of sgn : W (F ) → C X(F ), Z . We shall show
¡ that ¢it too is a
2-primary torsion group. This generalizes ¡the two observations
¢ that C X(F ), Z = 0 if F
is not formally real and sgn : W (F ) → C X(F ), Z is an isomorphism if F is euclidean.
We use and generalize the approach and results from [36] that we shall need in §41.
¡ If A ⊂¢ X(F ), write χA for the characteristic

¡ ¢ function of −1 A. In particular, χA ∈
C X(F ), Z if A is clopen. Let f ∈ C X(F ), Z . Then An = f (n) is a clopen set. As
{An | n ∈ Z} partition
P the compact space X(F ), only finitely many ¡ An are
¢ non-empty.
In particular, f = nχAn is a finite sum. This shows that C X(F ), Z is additively
generated by χA , as A varies over the clopen sets in the boolean space X(F ).
The finite intersections of the subbase elements (33.1)
(33.5) H(a1 , . . . , an ) := H(a1 ) ∩ · · · ∩ H(an ) with a1 , . . . , an ∈ F ×
form a base for the topology of X(F ). As
¡ ¢
H(a1 , . . . , an ) = supp hha1 , . . . , an ii ,
where supp b := {P ∈ X(F ) | sgnP b 6= 0} is the support of b, this base is none other than
the collection of clopen sets
(33.6) {supp(b) | b is a bilinear Pfister form}.
We also have
(33.7) sgn b = 2n χsupp(b)
if b is a bilinear n-fold Pfister form.
¡ ¢
Theorem 33.8. The cokernel of sgn : W (F ) → C X(F ), Z is 2-primary torsion.
Proof. It suffices to prove for each clopen set A ⊂ X(F ) that 2n χA ∈ im(sgn) for
some n ≥ 0. As X(F ) is compact, A is a finite union of clopen sets of the form (33.6)
whose characteristic functions lie in im(sgn) by (33.7). By induction, it suffices to show
that if A and B are clopen sets in X(F ) with 2n χA and 2m χB lying in im(sgn) for some
integers m and n then 2s χA∪B lies in im(sgn) for some s. But
(33.9) χA∪B = χA + χB − χA · χB ,
so
(33.10) 2m+n χA∪B = 2m (2n χA ) + 2n (2m χB ) − (2n χA ) · (2m χB )
lies in im(sgn) as needed. ¤
Define the reduced stability index str (F ) of F to be n if 2n is the exponent of the

cokernel of the signature map (or infinity if this exponent is not finite).
Refining the argument in the last theorem, we establish:
Lemma 33.11. Let C ⊂ X(F ) be clopen. Then there exists an integer n > 0 and a
b ∈ I n (F ) satisfying sgn b = 2n χC . More precisely, there exists an integer nP
> 0, bilinear
n-fold Pfister forms bi satisfying supp(bi ) ⊂ C, and integers ki such that ki sgn bi =
2n χC .
Proof. As X(F ) is compact and (33.6) is a base for the topology, there exists an
r ≥ 1 such that C = A1 ∪ · · · ∪ Ar with Ai = supp(bi ) for some mi -fold Pfister forms
bi , i = 1, . . . , r. We induct on r. If r = 1 the result follows by (33.7), so assume that
r > 1. Let A = A1 , b = b1 , and B = A2 ∪ · · · ∪ Ar . By induction, there exists an m ≥ 1
and a c ∈ I m (F ), a sum (and difference) of Pfister forms with the desired properties with
sgn c = 2m χB . Multiplying by a suitable power of 2, we may assume that m = m1 . Let
d = 2m (b ⊥ c) ⊥ (−b) ⊗ c. Then d is a sum (and difference) of Pfister forms whose
supports all lie in C as supp(a) = supp(2a) for any bilinear form a. By equations (33.9)
and (33.10), we have
22m χA∪B = 22m χA + 22m χB − 2m χA · 2m χB
= 2m (sgn b + sgn c) − sgn b · sgn c = sgn d,
so the result follows. ¤
Using the lemma, we can establish two useful results. The first is:
Theorem 33.12. (Normality Theorem) Let A and B be disjoint closed subsets of
X(F ). Then there exists an integer n > 0 and b ∈ I n (F ) satisfying
(
2n if P ∈ A
sgnP b =
0 if P ∈ B.
Proof. The complement X(F ) \ B is a union of clopen sets. As the closed set
A is covered by this union of clopen sets and X(F ) is compact, there exists a finite
cover¡ {C1 , .S
. . , Cr } of
¢ A for some clopen sets Ci , i = 1, . . . , r lying in X(F ) \ B. As
Ci \ Ci ∩ ( j6=i Cj ) is clopen for i = 1, . . . , r, we may assume this is a disjoint union.
By Lemma 33.11, there exist bi ∈ I m ), some mi , such that sgn bi = 2mi χCi . Let
P(F n−m
i
n = max{mi | 1 ≤ i ≤ r}. Then b = i 2 i

bi lies in I n (F ) and satisfies b = 2n χ∪i Ci .
Since A ⊂ ∪i Ci , the result follows. ¤
¡ ¢
We now investigate the relationship between elements in f ∈ C X(F ), 2m Z and
bilinear forms b satisfying 2m | sgnP b for all P ∈ X(F ). We first need a useful trick.
If ε = (ε1 , . . . , εn ) ∈ {±1}n and b = hha1 , . . . , an ii with ai ∈ F × , let
bε = hhε1 a1 , . . . , εn an ii.
Then supp(bε ) ∩ supp(bε0 ) = ∅ unless ε = ε0 .
Lemma
P 33.13. Let b be a bilinear n-fold Pfister form over an arbitrary field F . Then
2n h1i = ε bε in W (F ), where the sum runs over all ε ∈ {±1}n .
33. THE TOTAL SIGNATURE 137
Proof. Let b = hha1 , . . . , an ii and c = hha1 , . . . , an−1 ii with ai ∈ F × . As hh−1ii =

hhaii + hh−aii in W (F ) for all a ∈ F × , we have
X X X X
bε = cε0 hhan ii + cε0 hh−an ii = 2 cε0
ε ε0 ε0 ε0
where the ε0 run over all {±1}n−1 . The result follows by induction on n. ¤
The generalization of Lemma 33.13 given in the following exercise is useful. These
identities were first observed by Witt, who used them to give a simple proof that Wt (F )
is 2-primary (cf. Exercise 31.24(3)).
Exercise 33.14. Let a1 , . . . , an ∈ F × and ε = (ε1 , . . . , εn ) ∈ {±1}n . Suppose that
b = hha1 , . . . , an ii, c = h−a1 , . . . , −an i, and eε = hε1 , . . . , εn i. Then the following are true
in W (F ):
P
(1) c · bε = eε · bε = ( i εi )bε .
P
(2) 2n c = ε eε · bε .
Using Lemma 33.11, we also establish:
¡ ¢
Theorem 33.15. Let f ∈ C X(F ), 2m Z . Then there is a positive integer n and a
b ∈ I m+n (F ) such that 2n f =
Psgn b. More precisely, there exists an integer n such that
2n f can be written as a sum ri=1 ki sgn bi for some integers ki and bilinear (n + m)-fold
Pfister forms bi such that supp(bi ) ⊂ supp(f ) for every i = 1, . . . , r and whose supports
are pairwise disjoint.
Proof. We first show:
¡ ¢
Claim: Let g ∈ C X(F ), Z . Then Prthere exists a non-negative integer n and bilinear n-fold
n
Pfister forms ci such that 2 g = i=1 si sgn ci for some integers si with supp(ci ) ⊂ supp(g)
for every i = 1, . . . , r.
P
i iχg −1 (i) with i ∈ Z and each g (i) a
−1
The function g is a finite sum of functions
−1
clopen set. For each non-empty g (i), there exist a non-negative integer ni , bilinear
ni -fold Pfister forms bij with supp(bij ) ⊂ g −1 (i) and integers kj satisfying
P P 2ni χg−1 (i) =
n n−ni
j kj sgn bij by Lemma 33.11. Let n = maxi {ni }. Then 2 g = i,j ikj sgn(2 bij ).
This proves the Claim.
P
Let g = f /2m . By the Claim, 2n g = ri=1 si sgn ci for Pr some n-fold Pfister forms ci
whose supports lie in supp(g) = supp(f ). Thus 2 f = i=1 si sgn 2 ci with each 2m ci
n m
an (n + m)-fold Pfister form. P Let dm = c1 ⊗ · · · ⊗ ¡cr , an rn-fold

¢ Pfister form. By Lemma
33.13, we have 2 (n+1)r
f = ε sgn(2 si ci · dε ) in C X(F ), Z where ε runs over all {±1}rn .
For each i and ε, the form ci · dε is isometric to either 2n+m dε or is metabolic by Example
4.16(2) and (3). As the dε have pairwise disjoint suppports, adding the coefficients of the
isometric forms ci · dε yields the result. ¤
Corollary 33.16. Let b be a non-degenerate symmetric bilinear ¡form over F¢ and fix
m > 0. Then 2n b ∈ I n+m (F ) for some n ≥ 0 if and only if sgn b ∈ C X(F ), 2m Z .
Proof. We may assume that F is formally real as 2s(F )W (F ) = 0.
¡ ¢
⇒: ¡ If d is¢ a bilinear
¡ n-fold¢ Pfister form then sgn d ∈ C X(F ), 2n
Z . If follows that
sgn I n (F ) ¡⊂ C X(F ), 2n¢Z . Suppose that ¡ 2 n
b ∈ I n+m
¢ (F ) for some n ≥ 0. Then
n
2 sgn b ∈ C X(F ), 2 n+m
Z hence sgn b ∈ C X(F ), 2 Z .
m
⇐: By Theorem 33.15, there exists c ∈ I n+m (F ) such that sgn c = 2n sgn b. As Wt (F ) =

ker(sgn) is 2-primary torsion by the Local-Global Principle 31.25, there exists a non-
negative integer k such that 2n+k b = 2k c ∈ I n+m+k (F ). ¤
This Corollary 33.16 suggests that if b is a non-degenerate symmetric bilinear form

over F the following may be true
¡ ¢
(33.17) sgn b ∈ C X(F ), 2n Z if and only if b ∈ I n (F ) + Wt (F ).
This question was raised by Lam in [89].
In particular, in the case that F is a formally real pythagorean field, this would mean
b ∈ I n (F ) if and only if 2n | sgnP (b) for all P ∈ X(F )

as W (F ) is then torsion-free which would answer a question of Marshall in [98].
¡ ¢
Of course, if b ∈ I n (F ) + Wt (F ) then sgn b ∈ C X(F ), 2n Z . The converse would
follow if
2m b ∈ I n+m (F ) always implies that b ∈ I n (F ) + Wt (F ).
If F were formally real pythagorean the converse would follow if
2m b ∈ I n+m (F ) always implies that b ∈ I n (F ).
Because the nilradical of W (F ) is the torsion Wt (F ) when F is formally real, the total
signature induces an embedding of the reduced Witt ring
¡ ¢
Wred (F ) := W (F )/nil W (F ) = W (F )/Wt (F )
¡ ¢
into C X(F ), Z . Moreover, since Wt (F ) is 2-primary, the images of two non-degenerate
bilinear forms b and c are equal in the reduced Witt ring if and only if there exists a
non-negative integer n such that 2n b = 2n c in W (F ). Let : W (F ) → Wred (F ) be the
canonical ring epimorphism. Then the problem above becomes: If b is a non-degenerate
symmetric bilinear form over F then
¡ ¢
n
b ∈ Ired (F ) if and only if sgn b ∈ C X(F ), 2n Z .
n
where Ired (F ) is the image of I n (F ) in Wred (F ).
This is all, in fact, true as we shall see in §41 (Cf. Corollaries 41.9 and 41.10).
34. Bilinear and quadratic forms under quadratic extensions
In this section we develop the relationship between bilinear and quadratic forms over
a field F and over a quadratic extension K of F . We know that bilinear and quadratic
forms can become isotropic over a quadratic extension and we exploit this. We also
investigate the transfer map taking forms over K to forms over F induced by a nontrivial
F -linear functional. This leads to useful exact sequences of Witt rings and Witt groups
34. BILINEAR AND QUADRATIC FORMS UNDER QUADRATIC EXTENSIONS 139
done first in the case of characteristic not two in [39] and by Arason in [6] in the case of
characteristic two by Baeza in [17].
34.A. Bilinear forms under a quadratic extension. We start with bilinear forms
and the Witt ring.
Proposition 34.1. Let K/F be a quadratic field extension and s : K → F a nontrivial
F -linear functional satisfying s(1) = 0. Let c be an anisotropic bilinear from over K.
Then there exist bilinear forms b over F and a over K such that c ' bK ⊥ a with s∗ (a)
anisotropic.
Proof. We induct on dim c. Suppose that s∗ (c) is isotropic. It follows that there is
a b ∈ D(c) ∩ F , i.e., c ' hbi ⊥ c1 for some c1 . Applying the induction hypothesis to c1
completes the proof. ¤
We need the following generalization of Proposition 34.1.
Lemma 34.2. Let K/F be a quadratic extension of F and s : K → F a nontrivial
F -linear functional satisfying s(1) = 0. Let f be a bilinear anisotropic n-fold Pfister form
over F and c a non-degenerate bilinear form over K such that fK ⊗ c is anisotropic. Then
there exists a bilinear form b over F and a bilinear form a over K such that fK ⊗ c '
(f ⊗ b)K ⊥ fK ⊗ a with f ⊗ s∗ (a) anisotropic.
Proof. Let d = fK ⊗ c. We may assume that s∗ (d) is isotropic. Then there exists
a b ∈ D(d) ∩ F . If c ' ha1 , . . . , an i, there exist xi ∈ D(f e K ), not all zero satisfying
b = x1 a1 + · · · + xn an . Let yi = xi if xi 6= 0 and yi = 1 otherwise. Then
fK ⊗ c ' fK ⊗ hy1 a1 , . . . , yn an i ' fK ⊗ hb, z2 , . . . , zn i
for some zi ∈ K × as G(fK ) = D(fK ). The result follows easily by induction. ¤
Corollary 34.3. Let K/F be a quadratic extension of F and s : K → F a nontrivial
F -linear functional satisfying s(1) = 0. Let f be a bilinear anisotropic n-fold Pfister form
and c an anisotropic bilinear form over K satisfying f ⊗ s∗ (c) is hyperbolic. Then there
exists a bilinear form b over F such that dim b = dim c and fK ⊗ c ' (f ⊗ b)K .
Proof. If fK ⊗ c is anisotropic, the result follows by Lemma 34.2, so we may assume
that fK ⊗ c is isotropic. If fK is isotropic, it is hyperbolic and the result follows easily,
so we may assume the Pfister form fK is anisotropic. Using Proposition 6.22, we see
that there exist a bilinear form d with fK ⊗ d anisotropic and an integer n ≥ 0 with
dim d + 2n = dim c and fK ⊗ c ' fK ⊗ (d ⊥ nH). Replacing c by d, we reduce to the
anisotropic case. ¤
Note that if K/F is a quadratic extension and s, s0 : K → F are F -linear functionals
satisfying s(1) = 0 = s0 (1) with s nontrivial then s0∗ = as∗ for some a ∈ F .
Theorem 34.4. Let K/F be a quadratic field extension and s : K → F a nonzero
F -linear functional such that s(1) = 0. Then the sequence
rK/F ∗ s
W (F ) −−−→ W (K) −
→ W (F )
is exact.
¡ ¢
Proof. Let b ∈ F × then the binary form s∗ hbiK is isotropic hence metabolic.
Thus s∗ ◦ rK/F = 0. Let c ∈ W (K). By Proposition 34.1, there exists a decomposition
c ' bK ⊥ c1 with b a bilinear form over F and c1 a bilinear form over K satisfying s∗ (c1 )
is anisotropic. In particular, if s∗ (c) = 0, we have c = bK . This proves exactness. ¤
If K/F is a quadratic extension, denote the quadratic norm form of the quadratic
algebra K by NK/F . (Cf. §98.B.)
Lemma 34.5. Let K/F be a quadratic extension and s : K → F a nontrivial F -linear
functional. Let b be an anisotropic binary
¡ bilinear
¢ form over F such that the quadratic
form b ⊗ NK/F is isotropic. Then b ' s∗ hyi for some y ∈ K × .
Proof. Let {1, x} be a basis of K over F . Let c be the polar form of NK/F . We have
c(1, x) = NK/F (1 + x) − NK/F (x) − NK/F (1) = TrK/F (x)
for every x ∈ K. By assumption there are nonzero vectors v, w ∈ Vb such that
0 = (b ⊗ NK/F )(v ⊗ 1 + w ⊗ x)
= b(v, v) NK/F (1) + b(v, w)c(1, x) + b(w, w) NK/F (x)
= b(v, v) + b(v, w) TrK/F (x) + b(w, w) NK/F (x)
by the definition of tensor product (8.14). Let f : K → F be an F -linear functional
satisfying f (1) = b(w, w) and f (x) = b(v, w). By (98.2), we have
¡ ¢
f (x2 ) = f − TrK/F (x)x − NK/F (x) = − TrK/F (x)b(v, w) − NK/F (x)b(w, w) = b(v, v).
Therefore, the F
¡ -linear
¢ isomorphism K → Vb taking 1 to w and x to v is an isometry
between c = f∗ h1i and b. As f is the composition of s with the
¡ endomorphism
¢ ¡ ¢ of K
×
given by multiplication by some element y ∈ K , we have b ' f∗ h1i ' s∗ hyi . ¤
Proposition 34.6. Let K/F be a quadratic extension and s : K → F a nontrivial F -
linear functional. Let b be an anisotropic bilinear form over F . Then there exist bilinear
forms c over K and d over F such that b ' s∗ (c) ⊥ d and d ⊗ NK/F is anisotropic.
Proof. We induct on dim b. Suppose that b ⊗ NK/F is isotropic. Then there is a
2-dimensional
¡ ¢subspace W ⊂ Vb× with (b|W ) ⊗ NK/F isotropic. By Lemma 34.5, we have
b|W ' s∗ hyi for some y ∈ K . Applying the induction hypothesis to the orthogonal
complement of W in V completes the proof. ¤
√
Theorem 34.7. Let K = F ( a) be a quadratic field extension of F with a ∈ F × . Let
s : K → F be a nontrivial F -linear functional such that s(1) = 0. Then the sequence
∗s hhaii
W (K) −
→ W (F ) −−→ W (F )
is exact where the last homomorphism is multiplication by hhaii.
¡ ¢
Proof. For every c ∈ W (F ) we have hhaiis∗ (c) = s∗ hhaiiK c = 0 as hhaiiK = 0.
Therefore, the composition of the two homomorphisms in the sequence is trivial. Since
NK/F ' hhaiiq , the exactness of the sequence now follows from Proposition 34.6. ¤
34.B. Quadratic forms under a quadratic extension. We now turn to quadratic

forms.
Proposition 34.8. Let K/F be a separable quadratic field extension and ϕ an anisotropic
quadratic form over F . Then ϕ ' b ⊗ NK/F ⊥ ψ with b a non-degenerate symmetric bi-
linear form and ψ a quadratic form satisfying ψK is anisotropic.
Proof. Since K/F is separable, the binary form σ := NK/F is non-degenerate. As
F (σ) ' K, the statement follows from Corollary 22.12. ¤
Theorem 34.9. Let K/F be a separable quadratic field extension and s : K → F a
nonzero functional such that s(1) = 0. Then the sequence
rK/F ∗s NK/F ∗
rK/F s
W (F ) −−−→ W (K) −
→ W (F ) −−−→ Iq (F ) −−−→ Iq (K) −
→ Iq (F )
is exact where the middle homomorphism is multiplication by NK/F .
Proof. In view of Theorem 34.4 and Propositions 34.6 and 34.8, it suffices to prove
exactness at Iq (K). Let ϕ ∈ Iq (K) be an anisotropic form such that s∗ (ϕ) is hyperbolic.
We show by induction on n = dimK ϕ that ϕ ∈ im(rK/F ). We may assume that n > 0.
Let W ⊂ Vϕ be a totally isotropic F -subspace for the form s∗ (ϕ) of dimension n. As
ker(s) = F we have ϕ(W ) ⊂ F .
We claim that the K-space KW properly contains W , in particular,
1 1 n
(34.10) dimK KW = dimF KW > dimF W = .
2 2 2
To prove the claim choose an element x ∈ K such that x2 ∈ / F . Then for every nonzero
w ∈ W , we have ϕ(xw) = x2 ϕ(w) ∈ / F , hence xw ∈ KW but x ∈ / W . It follows
from the inequality (34.10) that the restriction of bϕ on KW and therefore on W is
nonzero. Consequently, there is a 2-dimensional F -subspace U ⊂ W such that bϕ |U is non-
degenerate. Therefore, the K-space KU is also 2-dimensional and the restriction ψ = ϕ|U
is a non-degenerate binary quadratic form over F satisfying ψK ' ϕ|KU . Applying the
induction hypothesis to (ψK )⊥ , we have (ψK )⊥ ∈ im(rK/F ). Therefore, ϕ = ψK +(ψK )⊥ ∈
im(rK/F ). ¤
¡ ¢
Remark 34.11. In Proposition 34.9, we have ker rK/F : Iq (F ) → Iq (K) = W (F )hha]]
when K = Fa .
Application of the above in the case of fields of characteristic not two provides a proof
of the following result shown in [39] and by Arason in [6]:
√
Corollary 34.12. Suppose that char F 6= 2 and K = F ( a)/F is a quadratic field
extension with a ∈ F × . If s : K → F is a nontrivial F -linear functional such that s(1) = 0
then the triangle
W (K)
: II
rK/F uu II s∗
uuu II
uu II
uu I$
W (F ) o W (F )
·hhaii
is exact.
Proof. Since the quadratic norm form NK/F coincides with ϕb where b = hhaii,
the map W (F ) → Iq (F ) given by multiplication by NK/F is identified with the map
W (F ) → I(F ) given by multiplication by hhaii. Note also that ker(rK/F ) ⊂ I(F ), so the
statement follows from Theorem 34.9. ¤
√
Remark 34.13. Suppose that char F 6= 2 and K = F ( a) is a quadratic extension
of F . Let b be an anisotropic bilinear form. Then by Proposition 34.8 and Example 9.5,
we see that the following are equivalent:
(1) bK is metabolic.
(2) b ∈ hhaiiW (F ).
(3) b ' hhaii ⊗ c for some symmetric bilinear form c.
In the case that char F = 2, Theorem 34.9 can be slightly improved.
We need the following computation:
Lemma 34.14. Let F be a field of characteristic 2 and K/F a quadratic field extension.
Let s : K → F be a nonzero F -linear functional satisfying s(1) = 0. Then for every x ∈ K
we have ½
¡ ¢ 0 if x ∈ F
s∗ hhx]] =
s(x)hhTrK/F (x)]] otherwise.
¡ ¢
In particular s∗ hhx]] ≡ hhTrK/F (x)]] modulo Iq2 (F ).
Proof. The element x satisfies the quadratic equation x2 + ax + b = 0 for some a, b ∈
F . We have TrK/F (x) = a and s(x2 ) = as(x) = s(x) TrK/F (x). Let x̄ = TrK/F (x) − x.
The element x̄ satisfies the same quadratic equation and s(x̄2 ) = s(x) TrK/F (x̄).
Let {v, w} be the standard basis for the space V of the form ϕ := hhx]] over K. If
x ∈ F then v and w span the totally isotropic F -subspace of s∗ (ϕ), i.e., s∗ (ϕ) = 0.
Suppose that x ∈ / F . Then V = W ⊥ W 0 with W = F v ⊕ F xw and W 0 = F x̄v ⊕ F w.
We have s∗ (ϕ) ' s∗ (ϕ)|W ⊥ s∗ (ϕ)|W 0 . As s∗ (ϕ)(v) = s(1) = 0, the form s∗ (ϕ)|W is
isotropic and therefore s∗ (ϕ)|W ' H. Moreover,
¡ ¢
s∗ (ϕ)(x̄v) = s(x̄2 ) = s(x) TrK/F (x), s∗ (ϕ)(w) = s(x) and s∗ bϕ (x̄v, w) = s(x̄) = s(x)
hence s∗ (ϕ)|W 0 ' s(x)hhTrK/F (x)]]. ¤
Corollary 34.15. Suppose that char F = 2. Let K/F be a separable quadratic field
extension and s : K → F a nonzero functional such that s(1) = 0. Then the sequence
rK/F ∗ s · NK/F rK/F ∗ s
0 → W (F ) −−−→ W (K) −
→ W (F ) −−−−→ Iq (F ) −−−→ Iq (K) −
→ Iq (F ) → 0
is exact.
Proof. To prove the injectivity of rK/F , it suffices to show that if b is an anisotropic
bilinear form over F then bK is also anisotropic. Let x ∈ K \ F be an element satisfying
x2 + x + a = 0 for some a ∈ F and let bK (v + xw, v + xw) = 0 for some v, w ∈ Vb . We
have
0 = bK (v + xw, v + xw) = b(v, v) + ab(w, w) + xb(w, w),
hence b(w, w) = 0 = b(v, v). Therefore v = w = 0 as b is anisotropic.
¡ ¢
By Lemma 34.14, we have for every y ∈ K, the form s∗ hhy]] is similar to hhTrK/F (y)]].
As the map s∗ is W (F )-linear, Iq (F ) is generated by the classes of binary forms and the
trace map TrK/F is surjective, the last homomorphism s∗ in the sequence is surjective. ¤
34.C. The filtrations of the Witt ring and Witt group under quadratic
extensions. We turn to the study of relations between the ideals I n (F ), I n (K) and
between the groups Iqn (F ) and Iqn (K) for a quadratic field extension K/F .
Lemma 34.16. Let K/F be a quadratic extension. Let n ≥ 1.
(1) We have
I n (K) = I n−1 (F )I(K),
i.e., I n (K) is the W (F )-module generated by n-fold bilinear Pfister forms bK ⊗ hhxii
with x ∈ K × and b an (n − 1)-fold bilinear Pfister form over F .
(2) If char F = 2 then
Iqn (K) = I n−1 (F )Iq (K) + I(K)Iqn−1 (F ).
Proof. (1): Clearly, to show that I n (K) = I n−1 (F )I(K), it suffices to show this for
the case n = 2. Let x, y ∈ K \ F . As 1, x, y are linearly dependent over F , there are
a, b ∈ F × such that ax + by = 1. Note that the form hhax, byii is isotropic and therefore
metabolic. Using the relation
hhuv, wii = hhu, wii + uhhv, wii
in W (K), we have
0 = hhax, byii = hhx, byii + ahhx, byii = hha, bii + bhha, yii + ahhx, bii + abhhx, yii,
hence hhx, yii ∈ I(F )I(K).
(2): In view of (1), it is sufficient to consider the case n = 2. The group Iq2 (K) is
generated by the classes of 2-fold Pfister forms by (9.6). Let x, y ∈ K. If x ∈ F then
hhx, y]] ∈ I(F )Iq (K). Otherwise y = a + bx for some a, b ∈ F . Then, by Lemma 15.1 and
Lemma 15.5,
hhx, y]] = hhx, a]] + hhx, bx]] = hhx, a]] + hhb, bx]] ∈ I(K)Iq (F ) + I(F )Iq (K)
since hhb, bx]] + hhx, bx]] = hhbx, bx]] = 0. ¤
Corollary 34.17. Let K/F be a quadratic extension and s : K → F a nonzero
F -linear functional. Then for every n ≥ 1:
¡ ¢
(1) s∗ ¡I n (K)¢ ⊂ I n (F ).
(2) s∗ Iqn (K) ⊂ Iqn (F ).
¡ ¢
Proof. (1): Clearly s∗ I(K) ⊂ I(F ). It follows from Lemma 34.16 and Frobenius
Reciprocity that
¡ ¢ ¡ ¢ ¡ ¢
s∗ I n (K) = s∗ I n−1 (F )I(K) = I n−1 (F )s∗ I(K) ⊂ I n−1 (F )I(F ) = I n (F ).
(2): This follows from (1) if char F 6= 2 and from Lemma 34.16(2) and Frobenius Reci-
procity if char F = 2. ¤
Lemma 34.18. Let K/F be a quadratic extension and s, s0 : K → F two nonzero

F -linear functionals. Let b ∈ I n (K). Then s∗ (b) ≡ s0∗ (b) mod I n+1 (F ).
Proof. As in the proof of Corollary 20.8, there exists a c ∈ K × such that s0∗ (c) =
s∗ (cc) for all symmetric bilinear¡forms c.¢ As b ∈ I n (K), we have hhcii · b ∈ I n+1 (K).
Consequently, s∗ (b) − s0∗ (b) = s∗ hhcii · b lies in I n+1 (F ). The result follows. ¤
Corollary 34.19. Let K/F ¡ be¢a quadratic field extension and s : K → F a nontrivial
F -linear functional. Then s∗ hhxii ≡ hhNK/F (x)ii modulo I 2 (F ) for every x ∈ K × .
¡ ¢
Proof. By Lemma 34.18, we know that s∗ hhxii is independent of the nontrivial
F -linear functional s modulo I 2 (F ). Using the functional defined in (20.9), the result
follows by Corollary 20.14. ¤
Let K/F be a separable quadratic field extension and let s : K → F be a nontrivial
F -linear functional such that s(1) = 0. It follows from Theorem 34.9 and Corollary 34.17
that we have a well-defined complex
rK/F s · NK/F rK/F s

(34.20) I n (F ) −−−→ I n (K) −
→∗
I n (F ) −−−−→ Iqn+1 (F ) −−−→ Iqn+1 (K) −
→∗
Iqn+1 (F )
and this induces (where by, abuse of notation, we label the maps in the same way)
n rK/F n ∗ s n · NK/F n+1 rK/F n+1 s∗ n+1
(34.21) I (F ) −−−→ I (K) −
→ I (F ) −−−−→ I q (F ) −−−→ I q (K) −
→ Iq (F ).
n
where I (F ) := I n (F )/I n+1 (F ).
By Lemma 34.18, it follows that the homomorphism s∗ in (34.21) is independent of
the nontrivial F -linear functional K → F although it is not independent in (34.20).
We show that the complexes (34.20) and (34.21) are exact on bilinear Pfister forms.
More precisely we have
Theorem 34.22. Let K/F be a separable quadratic field extension and s : K → F a
nontrivial F -linear functional such that s(1) = 0.
(1) Let c be an anisotropic bilinear n-fold Pfister form over K. If s∗ (c) ∈ I n+1 (F )
then there exists a bilinear n-fold Pfister form b over F such that c ' bK .
(2) Let b be an anisotropic bilinear n-fold Pfister form over F . If b·NK/F ∈ I n+2 (F ),
then there exists a bilinear n-fold Pfister form c over K such that b = s∗ (c).
(3) Let ϕ be an anisotropic quadratic (n + 1)-fold Pfister form over F . If rK/F (ϕ) ∈
I n+2 (K) then there exists a bilinear n-fold Pfister form b over F such that ϕ '
b ⊗ NK/F .
(4) Let ψ be an anisotropic (n + 1)-fold quadratic Pfister form over K. If s∗ (ψ) ∈
I n+2 (F ) then there exists a quadratic (n + 1)-fold Pfister form ϕ over F such that
ψ ' ϕK .
Proof. (1): As c represents 1, the form s∗ (c) is isotropic and belongs to I n+1 (F ). It
follows from the Hauptsatz 23.8 that s∗ (c) = 0 in W (F ). We show by induction on k ≥ 0
that there is a bilinear k-fold Pfister form d over F and a bilinear (n − k)-fold Pfister form
e over K such that c ' dK ⊗ e. The statement that we need follows when k = n.
Suppose we have d and e for some k < n. We have

0 = s∗ (c) = s∗ (dK · e0 ⊥ dK ) = s∗ (dK · e0 )
in W (F ) where, as usual, e0 denotes the pure subform of e. In particular, s∗ (dK ⊗ e0 ) is
isotropic. Thus there exists b ∈ F × ∩ D(dK ⊗ e0 ). It follows that dK ⊗ e ' dK ⊗ hhbii ⊗ f
for some Pfister form f over K by Theorem 6.15.
(2): By the Hauptsatz 23.8, we have b ⊗ NK/F is hyperbolic. We claim that b ' hhaii ⊗ a
for some a ∈ NK/F (K × ) and an (n − 1)-fold bilinear Pfister form a over F . If char F 6= 2,
the claim follows from Corollary 6.14. If char F = 2, it follows from Lemma 9.12 that
NK/F ' hha]] for some a ∈ D(b0 ). Clearly, a ∈ NK/F (K × ) and by Lemma 6.11, the form
b is divisible by hhaii. The claim is proven. ¡ ¢
× ×
¡ As a ∈¢ N K/F (K ) there is y ∈ K such that s ∗ hhyii = hhaii. It follows that
s∗ hhyii · a = hhaii · a = b.
(3): By the Hauptsatz 23.8, we have rK/F (ϕ) = 0 in Iq (K). The field K is isomorphic
to the function field of the 1-fold Pfister form NK/F . The statement now follows from
Corollary 23.7.
(4): In the case char F 6= 2, the statement follows from (1). So we may assume that
char F = 2. As ψ represents 1, the form s∗ (ψ) is isotropic and belongs to Iqn+2 (F ). It
follows from the Hauptsatz 23.8 that s∗ (ψ) = 0 ∈ Iq (F ). We show by induction for each
k = 1, . . . , n that there is a k-fold bilinear Pfister form d over F and a quadratic Pfister
form ρ over K such that ψ ' dK ⊗ ρ.
Suppose we have d and ρ for some k < n. As dim(dK ⊗ ρ0 ) > 12 dim(dK ⊗ ρ) with ρ0
the pure subform of ρ, the subspace of s∗ (dK ⊗ ρ0 ) intersects a totally isotropic subspace
of s∗ (dK ⊗ ρ) and therefore is isotropic. Hence there is a c ∈ F satisfying c ∈ D(dK ⊗ ρ) \
D(dK ). By Proposition 15.7, ψ ' d ⊗ hhciiK ⊗ µ for some quadratic Pfister form µ.
Applying the statement with k = n, we get¡an n-fold ¢ bilinear Pfister form b over F
such that ψ ' bK ⊗ hhy]] for some y ∈ K. As s∗ hhy]] is similar to hhTrK/F (y)]], we have
b ⊗ hhTrK/F (y)]] = 0 ∈ Iq (F ). By Corollary 6.14, TrK/F (y) = b + b2 + b0 (v, v) for some
b ∈ F and v ∈ Vb0 . Let x ∈ K \ F be an element such that x2 + x + a = 0 for some a ∈ F .
Set z = xb + (xb)2 + b0K (xv, xv) ∈ K and c = y + z. Since TrK/F (x) = TrK/F (x2 ) = 1, we
have TrK/F (z) = TrK/F (y). It follows that c ∈ F . Again by Corollary 6.14, we see that
bK ⊗ hhz]] is hyperbolic and therefore
¡ ¢
ψ = bK · hhy]] = bK · hhy + z]] = b · hhc]] K . ¤
√
of F . Let b be an anisotropic bilinear n-fold Pfister form over F . Then NK/F = hhaii so
by Theorem 34.22(3), the following are equivalent:
(1) bK ∈ I n+1 (K).
(2) b ∈ hhaiiW (F ).
(3) b ' hhaii ⊗ c for some (n − 1)-fold Pfister form c.
We now consider the case of a purely inseparable quadratic field extension K/F .
Lemma 34.24. Let K/F be a purely inseparable quadratic field extension and s : K → F
a nonzero F -linear functional satisfying s(1) = 0. Let b ∈ F × . Then the following condi-
tions are equivalent:
(1) b ∈ NK/F (K × ).
(2) hhbiiK = 0¡ ∈ W
¢ (K).
(3) hhbii = s∗ hyi for some y ∈ K × .
Proof. The equality NK/F (K × ) ¡= K¢ 2 ∩ F × proves (1) ⇔ (2). For any y ∈ F × ,
it follows by Corollary 34.19 that s∗ hyi is similar to hhNK/F (y)ii. This proves that
(1) ⇔ (3). ¤
Proposition 34.25. Let K/F be a purely inseparable quadratic field extension and
s : K → F a nontrivial F -linear functional such that s(1) = 0. Let b an anisotropic
bilinear form over F . Then there exist bilinear forms c over K and d over F satisfying
b ' s∗ (c) ⊥ d and dK is anisotropic.
Proof. We induct on dim b. Suppose that bK is isotropic. Then there is a 2-
dimensional
¡ subspace
¢ W ⊂ Vb such that (b|W )K is isotropic. By Lemma 34.24, we have
b|W ' s∗ hyi for some y ∈ K × . Applying the induction hypothesis to the orthogonal
complement of W in V completes the proof. ¤
Theorem 34.4 and Proposition 34.25 yield
Corollary 34.26. Let K/F be a purely inseparable quadratic field extension and
s : K → F a nonzero F -linear functional such that s(1) = 0. Then the sequence
rK/F ∗ s rK/F
W (F ) −−−→ W (K) −
→ W (F ) −−−→ W (K)
is exact.
Let K/F be a purely inseparable quadratic field extension and s : K → F a nonzero
linear functional such that s(1) = 0. It follows from Corollaries 34.17 and 34.26 that we
have well-defined complexes
rK/F s rK/F
(34.27) I n (F ) −−−→ I n (K) −
→∗
I n (F ) −−−→ I n (K)
and
n rK/F n ∗ s n rK/F n
(34.28) I (F ) −−−→ I (K) −
→ I (F ) −−−→ I (K).
As in the separable case, the homomorphism s∗ in (34.28) is independent of the non-
trivial F -linear functional K → F by Lemma 34.18 although it is not independent in
(34.27).
We show that the complexes (34.27) and (34.28) are exact on quadratic Pfister forms.
Theorem 34.29. Let K/F be a purely inseparable quadratic field extension and s :
K → F a nontrivial F -linear functional such that s(1) = 0.
(1) Let c be anisotropic n-fold bilinear Pfister form over K. If s∗ (c) ∈ I n+1 (F ) then
there exists an b over K such that c ' bK .
(2) Let b be anisotropic n-fold bilinear Pfister form over F . If bK ∈ I n+1 (K) then
there exists an n-fold bilinear Pfister form c such that b = s∗ (c).
Proof. (1): The proof is the same as in Theorem 34.22(1).
(2): By the Hauptsatz 23.8, we have bK = 0 ∈ W (K). In particular, bK is isotropic and
hence there is a 2-dimensional subspace W ⊂ Vb with b|W non-degenerate and isotropic
over K. Hence
¡ b|
¢ W is similar to hhbii for some b ∈ F × . As hhbiiK = 0, by Lemma 34.24
hhbii = s∗ hhyii for some y ∈ K × . By Corollary 6.17, we have b ' hhbii ⊗ d for some
bilinear Pfister form d. Finally,
¡ ¢ ¡ ¢
b = hhbii · d = s∗ hhyii · d = s∗ hhyii · d ∈ W (F ). ¤
We shall show in Theorems 40.3, 40.5, and 40.6 that the complexes 34.20, 34.21, 34.27
and 34.28 are exact for any n. Note that the exactness for small n (up to 2) can be shown
by elementary means.
34.D. Torsion in the Witt ring under a quadratic extension. We turn to the
transfer of the torsion ideal in the Witt ring of a quadratic extension to obtain results
found in [34]. We need the following lemma.
Lemma 34.30. Let K/F be a quadratic field extension of F and b a bilinear Pfister
form over F .
(1) If c is an anisotropic bilinear form over K such that bK ⊗ c is defined over F
then ¡there exists
¢ a form d over F¡ such that
¢ bK ⊗ c ' (b ⊗ d)K .
(2) rK/F W (F ) ∩ bK W (K) = rK/F bW (F ) .
Proof. (1): Let c = ha1 , . . . , an i. We induct on dim c = n. By hypothesis, there is a
c ∈ F × ∩ D(bK ⊗ c). Write c = a1 b1 + · · · + an bn with bi ∈ D(b e K ). Let ci = bi if bi 6= 0
and 1 if not. Then e := ha1 c1 , . . . , an cn i represents c so e ' hci ⊥ f. Since bi ∈ GK (b), we
have
bK ⊗ c ' bK ⊗ e ' bK ⊗ hci ⊥ bK ⊗ f.
As bK ⊗ f ∈ im(rK/F ), its anisotropic part is defined over F by Proposition 34.1 and
Theorem 34.4. By induction, there exists a form g such that bK ⊗ f ' bK ⊗ gK . Then
hci ⊥ g works.
(2) follows easily from (1). ¤
√
Proposition 34.31. Let K = F ( a)/F be a quadratic extension with a ∈ F × and
s : K → F a nontrivial F -linear functional such that s(1) = 0. Let b be an n-fold bilinear
Pfister form. Then
¡ ¢ ¡ ¢
s∗ W (K) ∩ annW (F ) (b) = s∗ annW (K) (bK ) .
Proof. By Frobenius Reciprocity, we have
¡ ¢ ¡ ¢
s∗ annW (K) (bK ) ⊂ s∗ W (K) ∩ annW (F ) (b).
¡ ¢
Conversely, if c ∈ s∗ W (K) ∩ annW (F ) (b), we can write c = s∗ (d) for some form d over
K. By Theorem 34.4 and Lemma 34.30,
¡ ¢ ¡ ¢
bK ⊗ d ∈ rK/F W (F ) ∩ bK W (K) = rK/F bW (F ) .
Hence there exists a form

¡ e over F such
¢ that bK ⊗ d = (b ⊗ e)K . Let f = d ⊥ −eK . Then
c = s∗ (d) = s∗ (f) ∈ s∗ annW (K) (bK ) as needed. ¤
The torsion Wt (F ) of W (F ) is 2-primary. Thus applying the proposition to ρ = 2n h1i
for all n yields
√
Corollary 34.32. Let K = F ( a) be a quadratic extension of F with a¡∈ F × and ¢
s :¡K → F¢ a nontrivial F -linear functional such that s(1) = 0. Then Wt (F )∩s∗ W (K) =
s∗ Wt (K) .
We also have the following:
Corollary
√ 34.33. Suppose that F is a field of characteristic different from two and
K = F ( a) a quadratic extension of F . Let s : K → F be a non-trivial F -linear
functional such that s(1) = 0. Then
¡ ¢ ¡ ¢
hhaiiW (F ) ∩ annW (F ) 2h1i = ker(rK/F ) ∩ s∗ W (K) ⊂
¡ ¢ ¡ ¢ ¡ ¢
annW (F ) 2h1i ∩ annW (F ) hhaii = s∗ annW (K) (2h1i) .
Proof. As hha, aii ' hh−1, aii, we have
¡ ¢ ¡ ¢
hhaiiW (F ) ∩ annW (F ) 2h1i = hhaiiW (F ) ∩ annW (F ) hhaii
¡ ¢
which yields the first equality by ¡Corollary
¢ 34.12. As
¡ hhaiiW
¢ (F¡) ⊂ ann W (F ) hhaii
¢ ,
we have the inclusion. Finally, s∗ W (K) ∩ annW (F ) 2h1i = s∗ annW (K) (2h1iK ) by
Proposition 34.31, so Corollary 34.12 yields the second equality. ¤
Remark 34.34. Suppose that F is a formally real field and K a quadratic extension.
Let s∗ : W (K) → W (F ) be the transfer induced by a nontrivial F -linear functional such
that s(1) = 0. Then it follows by Corollaries 34.12 and 34.32 that the maps induced by
rK/F and s∗ yield an exact sequence
rK/F ∗ s
0 → Wred (K/F ) → Wred (F ) −−−→ Wred (K) −→ Wred (F )
¡ ¢
(again abusing notation for the maps) where Wred (K/F ) := ker Wred (F ) → Wred (K) .
By Corollary 33.15, we have a zero sequence
rK/F s
n n n ∗ n
0 → Ired (K/F ) → Ired (F ) −−−→ Ired (K) −
→ Ired (F )
n
¡ n n
¢
where Ired (K/F ) := ker Ired (F ) → Ired (K) .
In fact, we shall see in §41 that this sequence is also exact.
35. Torsion in I n (F ) and torsion Pfister forms
In this section we study the property that I(F ) is nilpotent, i.e., that there exists an n
such that I n (F ) = 0. For such an n to exist, the field must be non-formally real. In order
to study all fields we broaden this investigation to the study of the existence of an n such
that I n (F ) is torsion-free. We wish to establish the relationship between this occurring
over F and over a quadratic field extension K. This more general case is more difficult,
so in this section we look at the simpler property that there are no torsion bilinear n-fold
Pfister forms over the field F . This was the approach introduced and used in [39] and
35. TORSION IN I n (F ) AND TORSION PFISTER FORMS 149
would be equivalent to I n (F ) being torsion-free if we knew that torsion bilinear n-fold

Pfister forms generate the torsion in I n (F ). This is in fact true as shown in [9] and which
we shall show in §41, but cannot be proven by these elementary methods.
35.A. The property An . In this section we begin the study torsion in I n (F ) for a
field F . We set
Itn (F ) := Wt (F ) ∩ I n (F ).
Note that the group It (F ) is generated by torsion binary forms by Proposition 31.30.
It is obvious that
Itn (F ) ⊃ I n−1 (F )It (F ).
Proposition 35.1. It2 (F ) = I(F )It (F ).
¡ ¢
Proof. Note that for all a, a0 ∈ F × and w, w0 ∈ D ∞h1i , we have
ahhwii + a0 hhw0 ii = ahh−aa0 , wii + a0 whhww0 ii,
hence
ahhwii + a0 hhw0 ii ≡ a0 whhww0 ii mod I(F )It (F ).
Let b ∈ It2 (F ). By
¡ Proposition
¢ 31.30, we have b is a sum of binary forms ahhwii with
a ∈ F × and w ∈ D ∞h1i . Repeated application of the congruence above shows that
b is congruent to a binary form ahhwii modulo I(F )It (F ). As ahhwii ∈ I 2 (F ) we have
ahhwii = 0 and therefore b ∈ I(F )It (F ). ¤
We shall prove in §41 that the equality Itn (F ) = I n−1 (F )It (F ) holds for every n.
It is easy to determine Pfister forms of order 2 (cf. Corollary 6.14).
Lemma 35.2. Let b be a bilinear n-fold Pfister form. Then
¡ 2b
¢ = 0 in W (F ) if and
only if either char F = 2 or b = hhwii ⊗ c for some w ∈ D 2h1i and c an (n − 1)-fold
Pfister form.
Proposition 35.3. Let F be a field and n ≥ 1 an integer. The following conditions
are equivalent.
(1) There are no n-fold Pfister forms of order 2 in W (F ).
(2) There are no anisotropic n-fold Pfister forms of finite order in W (F ).
(3) For every m ≥ n, there are no anisotropic m-fold Pfister forms of finite order in
W (F ).
Proof. The implications (3) ⇒ (2) ⇒ (1) are trivial.
(1) ⇒ (3): If char F = 2 the statement is clear as W (F ) is torsion. Assume that
char F 6= 2. Let 2k b = 0 in W (F ) for some k ≥ 1 and b an m-fold Pfister form with
m ≥ n. We show by induction on k¡ that¢b = 0 in W (F ). It follows from Lemma 35.2 that
2k−1 b ' hhwii ⊗ c for some w ∈ D 2h1i and a (k + m − 2)-fold Pfister form c. Let d be
an (n − 1)-fold Pfister form dividing c. Again by Lemma 35.2, the form 2hhwii · d is 0 in
W (F ), hence by assumption, hhwii · d = 0 in W (F ). It follows that 2k−1 b = hhwii · c = 0
in W (F ). By the induction hypothesis, b = 0 in W (F ). ¤
We say that a field F satisfies An if the equivalent conditions of Proposition 35.3 hold.
It follows from the definition that the condition An implies Am for every m ≥ n. It follows
from Proposition 31.11 that F satisfies A1 if and only if F is pythagorean.
If F is not formally real, the condition An is equivalent to I n (F ) = 0 as the group
W (F ) is torsion.
As the group It (F ) is generated by torsion binary forms, the property An implies that
n−1
I (F )It (F ) = 0.
Exercise 35.4. Suppose that F is a field of characteristic not two. If K is a quadratic
extension of F , let sK : K → F be a nontrivial F -linear functional such that sK (1) = 0.
Show the following are equivalent:
(1) F satisfies An+1 .
F ( w) ¡ √ ¢ ¡ ¢
√
(2) s∗ √ Pn (F ( w)) = Pn (F ) for every w ∈ D ∞h1i .
F ( w) ¡ n ¡√ ¢ ¡ ¢
(3) s∗ I (F w)) = I n (F ) for every w ∈ D ∞h1i .
We now study the property An under field extensions. The case of fields of character-
istic two is easy.
Lemma 35.5. Let K/F be a finite extension of fields of characteristic two. Then
I n (F ) = 0 if and only if I n (K) = 0.
Proof. The property I n (E) = 0 for a field E is equivalent to [E : E 2 ] < 2n by
Example 6.5. We have [K : F ] = [K 2 : F 2 ], as the Frobenius map K → K 2 given by
x → x2 is an isomorphism. Hence
(35.6) [K : K 2 ] = [K : F 2 ]/[K 2 : F 2 ] = [K : F 2 ]/[K : F ] = [F : F 2 ].
Thus we have I n (K) = 0 if and only if I n (F ) = 0. ¤
Let F0 be a formally real field satisfying A1 , i.e., a pythagorean field. Let Fn =
F0 ((t1 )) · · · ((tn )) be the iterated Laurent series field over F0 . Then Fn is also formally
real pythagorean
√ (cf. Example 31.8), hence Fn satisfies An for all n ≥ 1. However,
Kn = Fn ( −1) does not satisfy An as hht1 , . . . , tn ii is an anisotropic form over the non-
formally real field Kn . Thus the property An is not preserved under quadratic extensions.
Nevertheless, we have
√
Proposition 35.7. Suppose that F satisfies An . Let K = F ( a) be a quadratic
extension of F with a ∈ F × . Then K satisfies An if either of the following two conditions
hold:
¡ ¢
(i) a ∈ D ∞h1i .
(ii) Every bilinear n-fold Pfister form over F becomes metabolic over K.
Proof. If char F = 2 then I n (F ) = 0 hence I n (K)
¡ = 0¢ by Lemma 35.5. So we may
assume that char F 6= 2. Let y ∈ K × satisfy y ∈ D 2h1iK and let e be an (n − 1)-fold
Pfister form over K. By Lemma 35.2, it suffices to show that b := hhyii ⊗ e is trivial
in W (K). Let s∗ : W (K) → W (F ) be the transfer induced by a nontrivial F -linear
functional s(1) = 0.
We claim that s∗ (b) = 0. Suppose that n = 1. Then s∗ (b) ∈ It (F ) = 0. So we may

assume that n ≥ 2. As I n−1 (K) is generated by Pfister forms of the form hhzii ⊗ dK with
z ∈ K × and d an (n − 2)-fold Pfister form over F by Lemma 34.16, we may assume that
b = hhy, zii ⊗ dK .
¡ ¢
We have s∗ hhy, zii ∈ It2 (F ) = I(F )It (F ) by Proposition 35.1. So
¡ ¢ ¡ ¢
s∗ hhy, zii · dK = s∗ hhy, zii · d
lies in I n−1 (F )It (F ) which is trivial by An . The claim is proven.
It follows that b = cK for some n-fold Pfister form c over F by Theorem 34.22. Thus
we are done¡ if every ¢ n-fold Pfister form over F becomes hyperbolic over K. So assume
m
that a ∈ D ∞h1i . As b is torsion in W (K), there exists an m such that 2 b = 0 in
W (F ). Thus 2m cK is hyperbolic so 2m c is a sum of binary forms xhhay 2 + x2 ii in W (F )
for some x, y, z in F by Corollary 34.12. In particular, 2m c is torsion so trivial by An for
F . The result follows. ¤
Corollary 35.8. Suppose that I n (F ) = 0 (in particular F is not formally real). Let
K/F be a quadratic extension. Then I n (K) = 0.
In general, the above corollary does not hold if K/F is not quadratic. For example, let
F be the quadratic closure of the rationals, so√I(F ) = 0. There exist algebraic extensions
K of F such that I(K) 6= 0, e.g., K = F ( 3 2). It is true, however, that in this case
I 2 (K) = 0. It is still an unanswered question whether I 2 (K) = 0 when K/F is finite
and F is an arbitrary quadratically closed field, equivalently whether the cohomological
2-dimension of a quadratically closed field is at most one.
If I n (F ) is torsion-free then F satisfies An . Conversely, if F satisfies A1 , then I(F ) is
torsion-free by Proposition 31.11. If F satisfies A2 then it follows from Proposition 35.1
that I 2 (F ) is torsion-free as It (F ) is generated by torsion binary forms.
Proposition 35.9. A field F satisfies A3 if and only if I 3 (F ) is torsion-free.
Proof. The statement is obvious if F is not formally real, so we may assume that
char F 6= 2. Let b ∈ I 3 (F ) be a torsion element. By Proposition 35.1,
r
X
b= xi hhyi , wi ii
i=1
¡ ¢
for some xi , yi ∈ F × and wi ∈ D ∞h1i . We show by induction on r that b = 0.
√
It follows from Proposition 35.7 that K = F ( w) with w = wr satisfies A3 . By the
induction hypothesis, we have bK = 0. Thus b = hhwii · c for some c ∈ W (F ) by Corollary
34.12. Then c must be even dimensional as the determinant of c is trivial. Choose d ∈ F ×
such that d := c + hhdii ∈ I 2 (F ).
Thus in W (F ),
b = hhwii · d − hhw, dii.
Note that hhwii · d = 0 in W (F ) by A3 . Consequently, hhw, dii ∈ I 3 (F ), so it is zero in
W (F ) by the Hauptsatz 23.8. This shows b = 0. ¤
We shall show in Corollary 41.5 below that I n (F ) is torsion-free if and only if F

satisfies An for every n ≥ 1.
We have an application for quadratic forms first shown in [38].
Theorem 35.10. (Classification Theorem) Let F be a field.
(1) Dimension and total signature classify the isometry classes of non-degenerate quadratic
forms over F if and only if Iq (F ) is torsion-free, i.e., F is pythagorean. In particular, if
F is not formally real then dimension classifies the isometry classes of forms over F if
and only if F is quadratically closed.
(2) Dimension, discriminant and total signature classify the isometry classes of non-
degenerate even dimensional quadratic forms over F if and only if Iq2 (F ) is torsion-free. In
particular, if F is not formally real then dimension and discriminant classify the isometry
classes of forms over F if and only if Iq2 (F ) = 0.
(3) Dimension, discriminant, Clifford invariant, and total signature classify the isometry
classes of non-degenerate even dimensional quadratic forms over F if and only if Iq3 (F )
is torsion-free. In particular, if F is not formally real then dimension, discriminant, and
Clifford invariant classify the isometry classes of forms over F if and only if Iq3 (F ) = 0.
Proof. We prove (3) as the others are similar (and easier). If Iq3 (F ) is not torsion-
free, then there exists an anisotropic torsion form ϕ ∈ P3 (F ) by Proposition 35.9 if F is
formally real and trivially if F is not formally real as then Iq (F ) is torsion. As ϕ and 4H
have the same dimension, discriminant, Clifford invariant, and total signature but are not
isometric, these invariants do not classify.
Conversely, assume that Iq3 (F ) is torsion-free. Let ϕ and ψ be non-degenerate even-
dimensional quadratic forms having the same dimension, discriminant, Clifford invariant,
and total signature. Then by Theorem 13.7, the form θ := ϕ ⊥ −ψ lies in Iq2 (F ) and is
torsion. As ϕ and ψ have the same dimension, it suffices to show that θ is hyperbolic.
Thus the result is equivalent to showing:
If a torsion form θ ∈ Iq2 (F ) has trivial Clifford invariant and Iq3 (F ) is torsion-free
then θ is hyperbolic.
The case char F = 2 follows from Theorem
Pr 16.3. So we may assume that char F 6= 2.
By Proposition 35.1, we can write θ = i=1 ai hhbi , ci ii in Iq (F ) with hhci ii torsion forms.
We prove that θ is hyperbolic by induction on r.
√
Let K = F ( c) with c = cr . Clearly, θK ∈ Iq2 (K) is torsion and has trivial Clifford
invariant. By Proposition 35.7 and Corollary 35.9, we have Iq3 (K) is torsion-free. By the
induction hypothesis, θK is hyperbolic. By Corollary 23.7, we conclude that θ = ψ · hhcii
in Iq (F ) for some quadratic form ψ. As disc(θ) is trivial, dim ψ is even. Choose d ∈ F ×
such that τ := ψ + hhdii ∈ I 2 (F ). Then
θ = τ · hhcii − hhd, cii
in W (F ).
As the torsion form τ ⊗ hhcii belongs to Iq3 (F ), it is hyperbolic. As the Clifford
invariant of θ is trivial, it follows that the Clifford invariant of hhd, cii must also be
trivial. By Corollary 12.5, hhd, cii is hyperbolic and hence θ is hyperbolic. ¤
Remark 35.11. The Stiefel-Whitney classes introduced in (5.4) are defined on non-
degenerate bilinear forms. If b is such a form then the wi (b) determine sgn b for every
P ∈ X(F ) by Remark 5.8 and Example 5.13. We also have wi = ei for i = 1, 2 by
Corollary 5.9.
Let b and b0 be two non-degenerate ¡ symmetric
¢ bilinear forms of the same dimension.
Suppose that w(b) = w(b0 ), then w [b] − [b0 ] = 1, where [ ] is the class of a form in
the Witt-Grothendieck ring W c (F ). It follows that [b] − [b0 ] lies in Ib3 (F ) by (5.11) hence
b − b0 lies in I 3 (F ). As the wi determine the total signature of a form, we have b − b0
is torsion by the Local-Global Principle 31.25. It follows that the dimension and total
Stiefel-Whitney class determine the isometry class of anisotropic bilinear forms if and only
if I 3 (F ) is torsion-free.
Suppose that char F 6= 2. Then all metabolic forms are hyperbolic, so in this case the
dimension and total Stiefel-Whitney class determine the isometry class of non-degenerate
symmetric bilinear forms if and only if I 3 (F ) is torsion-free. In addition, we can define
another Stiefel-Whitney map
c (F ) → H ∗ (F )[[t]]×
ŵ : W
to be the composition of w and the map k∗ (F )[[t]] → H ∗ (F )[[t]] induced by the norm
residue homomorphism h∗F : k∗ (F ) → H ∗ (F ) in §101.5. Then dimension and ŵ classify
the isometry classes of non-degenerate bilinear forms if and only if I 3 (F ) is torsion-free
by Theorem 35.10 as h∗ is an isomorphism if F is a real closed field and w̃2 is the classical
Hasse invariant so determines the Clifford invariant.
We turn to the question on whether the property An goes down. This was first done
in [39]. We use the proof given in [34].
Theorem 35.12. Let K/F be a finite normal extension. If K satisfies An so does F .
Proof. Let G = Gal(K/F ) and let H be a Sylow 2-subgroup of G. Set E = K H ,
L = K G . The field extension L/F is purely inseparable, so [L : F ] is either odd or L/F
is a tower of successive quadratic extensions. The extension K/E is a tower of successive
quadratic extensions and [E : L] is odd. Thus we may assume that [K : F ] is either 2 or
odd. Springer’s Theorem 18.5 solves the case of odd degree. Hence we may assume that
K/F is a quadratic extension.
The case char F = 2 follows from Lemma 35.5. Thus√we may assume that the charac-
teristic of F is different from two and therefore K = F ( a) with a ∈ F × . Let s : K → F
be a nontrivial F -linear functional with s(1) = 0.
Let b be a 2-torsion bilinear n-fold Pfister form.
¡ ¢We must show that b = 0 in W (F ).
As bK = 0 we have b ∈ hhaiiW (F ) ∩ annW (F ) 2h1i by Corollary 34.12. As hha, aii =
hha, −1ii, it follows that hhaii · b = 0 in W (F ) hence by ¡Corollary
¢ 6.14, we can write
b ' hhbii
¡ ¢⊗ c for some (n − 1)-fold Pfister form c and b ∈ D hhaii . Choose x ∈ K × such
that s∗ hxi = hhbii and let d = xcK . Then
¡ ¢
s∗ (d) = s∗ hxi c = hhbiic = b.
If d = 0 then b = 0 and we are done. So we may assume that d and therefore cK is

anisotropic.
We have s∗ (2d) = 2b = 0 in W (F ), hence the form s∗ (2d) is isotropic. Therefore, 2d
e K ) satisfying x(u + v) = c. But
represents an element c ∈ F × hence there exist u, v ∈ D(c
the form hhu+vii⊗c is 2-torsion and K satisfies An . Consequently, u+v ∈ D(cK ) = G(cK )
as cK is anisotropic. We have
d = xcK ' x(u + v)cK = ccK .
Therefore, 0 = s∗ (d) = b in W (F ) as needed. ¤
Corollary 35.13. Let K/F be a finite normal extension with F not formally real.
If I n (K) = 0 for some n then I n (F ) = 0.
The above corollary is, in fact, true without the condition that K/F be normal. We
have already shown this to be the case for fields of characteristic two in Lemma 35.5
and will show it to be true for characteristic not two in Corollary 43.9 below. In fact, it
follows from §43 below that if K/F a finitely generated field extension of transcendence
degree m then I n (K) = 0 implies n ≥ m and I n−m (F ) = 0. But we cannot do this by the
elementary means employed here.
Corollary 35.14. Let K/F be a quadratic extension.
(1) Suppose that I n (K) = 0. Then L satisfies An for every extension L/F such that
[L : F ] ≤ 2. ¡ ¢
(2) Suppose that I n (K) = 0. Then I n (F ) = hh−wiiI n−1 (F ) for every w ∈ D ∞h1i .
(3) Suppose that I n (F ) = hh−wiiI n−1 (F ) for ¡some ¢w ∈ F × . Then both F and K
satisfy An+1 and if char F 6= 2 then w ∈ D ∞h1i .
Proof. (1), (2): By Corollaries 35.8 and 35.13 if F is not formally real then I n (F ) = 0
if and only if I n (L) = 0 for any quadratic extension L/F . In particular (1) and (2) follow
if F is √not formally real. So suppose ¡ √that¢ F is formally real. We may nassume that
× n
K = F ( a) with a ∈ F . Then I L( a) = 0 by Proposition 35.7 hence I (L) satisfies
An by Theorem 35.12. This establishes (1).
¡ ¢ √ √
Let w ∈ D ∞h1i¡ . Then F¢ ( −w) is not formally real. By (1), the field F ( −w)
√
satisfies An hence I n F ( −w) = 0. In particular, if b is a bilinear n-fold Pfister form
then bF (√−w) is metabolic. Thus b ' hh−wii ⊗ c for some (n − 1)-fold Pfister form c over
F by Remark 34.23 and (2) follows.
(3): If char F = 2 then I n (F ) = 0 hence I n (K) = 0 by Corollary 35.8. So we may
assume that char F 6= 2. By Remark 34.23, we have 2n h1i ' hh−wii ⊗ b¡for some ¢ bilinear
(n − 1)-fold n e
¡ ¢Pfister form b. As 2 h1i only represents elements in D ∞h1i , we have
w ∈ D ∞h1i .
√
To show the first statement, it suffices to show that L = F ( −w) satisfies An+1 by (1)
and (2). Since I n+1 (L) is generated by Pfister forms of the type hhxii ⊗ cL where x ∈ L×
and c is an n-fold Pfister form over F by Lemma 34.16, we have I n+1 (L) ⊂ hh−wiiI n (L) =
{0}. ¤
If F is the field of 2-adic numbers then I 2 (F ) = 2I(F ) and K satisfies I 3 (K) = 0 for
all finite extensions K/F but no such K satisfies I 2 (K) = 0. In particular, statement (3)
of Corollary 35.14 is the best possible.
Corollary 35.15. ¡ Let¢ F be¡a field¢ extension of transcendence degree n over a real
closed field. Then D 2n h1i = D ∞h1i .
√ ¡ √ ¢
Proof. As F ( −1) is a Cn -field by Theorem 97.7 below, we have I n F ( −1) = 0.
Therefore, F satisfies An by Corollary 35.14. ¤
¡ √ ¢
35.B. Torsion-freeness and I n F ( −1) = 0. We turn to prove to a result of
Krüskemper
¡ 35.26¢(cf. [88]) showing F satisfying An and I n (F ) torsion free are equivalent
√
when I n F ( −1) = 0. We follow Arason’s notes (cf. [5]) generalizing our congruence
relations of Pfister forms developed in §24.
Let b be a bilinear Pfister form. For simplicity, we set
Ib (F ) = {c ∈ I(F ) | b · c = 0 ∈ W (F )} = I(F ) ∩ annW (F ) (b) ⊂ I(F ).
We note if b is metabolic then Ib (F ) = I(F ). We tacitly assume that b is anisotropic
below.
Lemma 35.16. Let c be a bilinear (n − 1)-fold Pfister form and d ∈ DF (b ⊗ c). Then
hhdii · c ∈ I n−1 (F )Ib (F ).
Proof. We induct on n. The hypothesis implies that b · hhdii · c = 0 in W (F ) hence
h1, −di · c ∈ Ib (F ). In particular, the case n = 1 is trivial. So assume that n > 1 and
that the lemma holds for (n − 2)-fold Pfister forms. Write c = hhaii ⊗ d where d is an
e ⊗ d). If e2 = 0 then we
(n − 2)-fold Pfister form. Then d = e1 − ae2 , where e1 , e2 ∈ D(b
are done by the induction hypothesis. So assume that e2 6= 0. Then d = e2 (e − a), where
e ⊗ d). By the induction hypothesis, we have
e = e1 /e2 ∈ D(b
hhdii · c = hhe2 (e − a)ii · c = hhe − aii · c + hhe − a, e2 ii · c
≡ hhe − aii · c mod I n−1 (F )Ib (F ).
It follows that we may assume that e2 = 1, hence that d = e − a. But then
hhd, aii = hhe − a, aii = hhe, a0 ii
for some a0 6= 0 by Lemma 4.15 , hence
hhdii · c = hhd, aii · d = hhe, a0 ii · d.
By the induction hypothesis, it follows that hhdii · c ∈ I n−1 (F )Ib (F ). ¤
Lemma 35.17. Let e be a bilinear n-fold Pfister form and b ∈ D(b ⊗ e0 ). Then there
is a bilinear (n − 1)-fold Pfister form f such that e ≡ hhbii · f mod I n−1 (F )Ib (F ).
Proof. We induct on n. If n = 1 then e0 = hhaii and b = ax for some x ∈ D(−b). It
follows that
hhbii = hhaxii = hhaii + ahhxii ≡ hhaii mod Ib (F ).
Now assume that n > 1 and that the lemma holds for (n − 1)-fold Pfister forms. Write
e ⊗ d0 )
e = hhaii ⊗ d with d an (n − 1)-fold Pfister form. Then b = c + ad, where c ∈ D(b
e ⊗ d). If d = 0 then we are done by the induction hypothesis. So assume that

and d ∈ D(b
d 6= 0. Then
hhadii · d = hhaii · d + ahhdii · d
≡ hhaii · d mod I n−1 (F )Ib (F )
by Lemma 35.16. It follows that we may assume that d = 1, hence b = c + a. If c = 0 then
b = a and there is nothing to prove. So assume that c 6= 0. By the induction hypothesis,
we can write
d ≡ hhcii · g mod I n−2 (F )Ib (F )
with g an (n − 2)-fold Pfister form. As
hha, cii = hhb − c, cii ' hhb, c0 ii
for some c0 6= 0 by Lemma 4.15, it follows that
e = hhaii · d ≡ hha, cii ⊗ g
= hhb, c0 ii · g mod I n−1 (F )Ib (F )
as needed. ¤
Lemma 35.18. Let e be a bilinear n-fold Pfister form and h a bilinear form over F .
(1) If e ∈ Ib (F) then e ∈ I n−1 (F )Ib (F ).
(2) If h · e ∈ Ib (F ) then h · e ∈ I n−1 (F )Ib (F ).
Proof. (1): The hypothesis implies that b · e = 0 in W (F ). In particular, b ⊗ e =
b ⊥ b ⊗ e0 is isotropic. It follows that there exists an element b ∈ DF (b) ∩ DF (b ⊗ e0 ). By
Lemma 35.17,
e ≡ hhbii · f ≡ 0 mod I n−1 (F )Ib (F ).
(2): The hypothesis implies that h · b · e = 0 in W (F ). If b · e = 0 in W (F ) then, by (1),
we have e ∈ I n−1 (F )Ib (F ) and we are done. Else we have h ∈ Ib⊗e (F ), which is generated
by the forms hhxii, with x ∈ D(b ⊗ e). It therefore suffices to prove the claim in the case
h = hhxii. But then, by (1), we even have h · e ∈ I n (F )Ib (F ). ¤
Lemma 35.19. Suppose the bilinear n-fold Pfister forms e, f satisfy
ae ≡ bf mod Ib (F )
with a, b ∈ F × . Then
ae ≡ bf mod I n−1 (F )Ib (F ).
Proof. We induct on n. As the case n = 1 is trivial, we may assume that n > 1
and that the claim holds for (n − 1)-fold Pfister forms. The hypothesis implies that
ab ⊗ e ' bb ⊗ f, in particular, b/a ∈ DF (b ⊗ e). By Lemma 35.16, we therefore have
ae ≡ bf mod I n−1 (F )Ib (F )
(actually, mod I n (F )Ib (F )). Hence we may assume that a = b. Dividing by a, we may
even assume that a = b = 1. Write
e = hhcii ⊗ g and f = hhdii ⊗ h
with g, h being (n−1)-fold Pfister forms. The hypothesis now implies that b⊗e0 ' b⊗f0 . In
particular, d ∈ D(b⊗e0 ). By Lemma 35.17, we can write e ≡ hhdii·g1 mod I n−1 (F )Ib (F )
with g1 an (n − 1)-fold Pfister form. It follows that we may assume that c = d. By the
induction hypothesis, we then have g ≡ h mod I n−2 (F )Ib⊗hhdii (F ), hence
hhdii · g ≡ hhdii · h mod hhdiiI n−2 (F )Ib⊗hhdii (F ).
We are therefore finished if we can show ¡that hhdiiI¢b⊗hhdii (F ) ⊆ I(F )Ib (F ). Now, Ib⊗hhdii (F )
is generated by the hhxii, with x ∈ D b ⊗ hhdii . For such a generator hhxii, we have
b · hhd, xii = 0 in W (F ), hence, by Lemma 35.18, the form hhd, xii lies in I(F )Ib (F ). ¤
Proposition 35.20. Let e, f, g be bilinear n-fold Pfister forms. Assume that
ae ≡ bf + cg mod Ib (F ).
Then
ae ≡ bf + cg mod I n−1 (F )Ib (F ).
Proof. The hypothesis implies that ab · e = bb · f + cb · g in W (F ). In particular, the
form bb ⊗ f ⊥ cb ⊗ g is isotropic. It follows that there exists d ∈ D(bb ⊗ f) ∩ D(−cb ⊗ g).
By Lemma 35.16, we then have
bf ≡ df mod I n−1 (F )Ib (F ) and cg ≡ −dg mod I n−1 (F )Ib (F )
(actually, mod I n (F )Ib (F )). Hence we may assume that c = −b. Dividing by b, we may
even assume that b = 1 and c = −1. Then the hypothesis implies that ab · e = b · f − b · g
in W (F ) and we have to prove that ae ≡ f − g mod I n−1 (F )Ib (F ).
As ab·e = b·f−b·g in W (F ), it follows that b⊗f and b⊗g are linked using Proposition
6.21 and with b dividing the linkage. Hence there exists an (n − 1)-fold Pfister form d
and elements b0 , c0 6= 0 such that b ⊗ f ' b ⊗ d ⊗ hhb0 ii and b ⊗ g ' b ⊗ d ⊗ hhc0 ii (and
hence b ⊗ e ' b ⊗ d ⊗ hhb0 c0 ii). By Lemma 35.19, we then have
f ≡ d · hhb0 ii and also g ≡ d · hhc0 ii mod I n−1 (F )Ib (F ).
We may therefore assume that f = d ⊗ hhb0 ii and g = d ⊗ hhc0 ii. Then f − g = d · h−b0 , c0 i =
−b0 d · hhb0 c0 ii in W (F ). The result now follows from Lemma 35.19. ¤
Remark 35.21. From Lemmas 35.16 – 35.19 and Proposition 35.20, we easily see that
the corresponding results hold for the torsion part It (F ) of I(F ) instead of Ib (F ). Indeed,
in each case, we only have to use our result for b = 2k h1i for some k ≥ 0.
We always have 2I n (F ) ⊂ I n+1 (F ) for a field F . For some interesting fields, we have
equality, i.e., 2I n (F ) = I n+1 (F ) for some positive integer n. In particular, we shall see in
Lemma 41.1 below this is true for some n for any field of finite transcendence degree over
its prime field. (This is easy if the field has positive characteristic but depends on Fact
16.2 when the characteristic of F is zero.) We shall now investigate when this phenomenon
holds for a field.
Proposition 35.22. Let F be a field. Then 2I n (F ) = I n+1 (F ) if and only if every
anisotropic bilinear (n + 1)-fold Pfister form b is divisible by 2h1i, i.e., b ' 2c for some
n-fold Pfister form c.
Proof. If 2h1i is metabolic, the result is trivial so assume not. In particular, we

may assume that char F 6= 2. Suppose 2I n (F ) = I n+1 (F ) and b is an anisotropic bilinear
(n+1)-fold Pfister form. By assumption, there exist d ∈ I n (F ) such that b = 2d in W (F ).
By Remark 34.23, we have b ' 2c for some n-fold Pfister form c. ¤
It is also useful to study a variant of the property that 2I n (F ) = I n+1 (F ) intro-

n
duced by Bröcker in [21]. Recall that Ired (F ) is the image of I n (F ) under the canon-
ical homomorphism W (F ) → Wred (F ) = W (F )/Wt (F ). We investigate the case that
n n+1
2Ired (F ) = Ired (F ) for some positive integer n. Of course, if 2I n (F ) = I n+1 (F ) then
n n+1
2Ired (F ) = Ired (F ). We shall show that the above proposition generalizes. Further, we
shall show this property is characterized by the cokernel of the signature map
¡ ¢
sgn : W (F ) → C X(F ), Z .
Recall that the cokernel of this map is a 2-primary group by Theorem 33.8.
Proposition 35.23. Suppose the reduced stability str (F ) is finite and equals n. Then
n n+1
n is the least non-negative integer such that 2Ired (F ) = Ired (F ). Moreover, for any
bilinear (n + 1)-fold Pfister form b, there exists an n-fold Pfister form c such that b ≡ 2c
mod Wt (F ).
¡ Proof. Let¢ b be an anisotropic bilinear (n+1)-fold Pfister form. In particular, sgn b ∈

C X(F ), 2 Z . By assumption, there exists a bilinear form d satisfying sgn d = 12 sgn b.
n+1
Thus b − 2d ∈ Wt (F ) hence there exists an integer m such that 2m b = 2m+1 d in W (F ) by

Theorem 31.21. If 2m b is metabolic the result is trivial, so we may assume it is anisotropic.
By Proposition 6.22, there exists f such that 2m b ' 2m+1 f. Therefore, 2m b ' 2m+1 c for
n n+1
some bilinear n-fold Pfister form c by Corollary 6.17. Hence 2Ired (F ) = Ired (F ).
¡ ¢
Conversely, suppose that 2Ired (F ) = Ired (F ). Let f ∈ C X(F ), Z . It suffices to
n n+1
show that there exists a bilinear form b satisfying sgn b = 2n f . By Theorem 33.15, there
exists an integer m and a bilinear form b ∈ I m (F ) satisfying sgn b = 2m f . So we are
done if m ≤ n. If m > n then there exists c ∈ I n (F ) such that sgn b = sgn 2m−n c and
2n f = sgn c. ¤
n n+1
Remark 35.24. If 2Ired (F ) = Ired (F ) then for any bilinear (n + m)-fold Pfister form
n+m
b there exists an n-fold Pfister form c such that b ≡ 2m c mod It (F ) and Ired (F ) =
m n n n+1
2 Ired (F ). Similarly, if 2I (F ) = I (F ) then for any bilinear (n + m)-fold Pfister form
b there exists an n-fold Pfister form c such that b ' 2m c and I n+m (F ) = 2m I n (F ).
n n+1
Suppose that 2Ired (F ) = Ired (F ). Let b be a n-fold Pfister form over F and let
d ∈ F × . Write
hhdii · b ≡ 2e mod It (F ) and hh−dii · b ≡ 2f mod It (F )
for some n-fold Pfister forms e and f over F . Adding, we then get 2b ≡ 2e+2f mod It (F ),
hence also b ≡ e + f mod It (F ). By Proposition 35.20, it follows that we even have
b ≡ e + f mod I n−1 (F )It (F ).
We generalize this as follows:
n n+1
Lemma 35.25. Suppose that 2Ired (F ) = Ired (F ). Let b be a bilinear n-fold Pfister
form and let d1 , . . . , dm ∈ F × . Write
hhε1 d1 , . . . , εm dm ii · b ≡ 2m cε mod It (F )
with cε a bilinear n-fold Pfister form for every ε = (ε1 , . . . , εm ) ∈ {±1}m . Then
X
b≡ cε mod I n−1 (F )It (F ).
ε
Proof. We induct on m. The case m = 1 is done above. So assume that m > 1.

Write hhε2 d2 , . . . , εm dm ii·b ≡ 2m−1 dε0 mod It (F ) with dε0 a bilinear n-fold Pfister form
P for
every ε0 = (ε2 , . . . , εm ) ∈ {±1}m−1 . By the induction hypothesis, we then have b ≡ ε0 dε0
mod I n−1 (F )It (F ). It therefore suffices to show that
dε0 ≡ c(+1,ε0 ) + c(−1,ε0 ) mod I n−1 (F )It (F )
for every ε0 . Since
¡ ¢
2m dε0 ≡ 2hhε2 d2 , . . . , εm dm ii · c = hhdii + hh−dii · hhε2 d2 , . . . , εm dm ii · e
≡ 2m c(+1,ε0 ) + 2m c(−1,ε0 ) mod It (F )
in W (F ), hence also dε0 ≡ c(+1,ε0 ) + c(−1,ε0 ) mod It (F ). By Proposition 35.20, it follows
that dε0 ≡ c(+1,ε0 ) + c(−1,ε0 ) mod I n−1 (F )It (F ). ¤
Krüskemper’s main result in [88] is:

n n+1
Theorem 35.26. Let 2Ired (F ) = Ired (F ). Then
Itn (F ) = I n−1 (F )It (F ).
P
Proof. Suppose that ri=1 ai bi ∈ It (F ), where b1 , . . . , br are bilinear n-fold
Pr Pfister
×
forms and ai ∈ F . We prove by induction on r that this implies that i=1 ai bi ∈
I n−1 (F )It (F ). The case r = 1 is simply Lemma 35.18, so assume that r > 1.
Write bi = hhai1 , . . . , ain ii for i = 1, . . . , r and let m = rn and
(d1 , . . . , dm ) = (a11 , . . . , a1n , a21 , . . . , a2n , . . . , ar1 , . . . , arn ).
Write
hhε1 d1 , . . . , εm dm ii · bi ≡ 2m ciε mod It (F )
with ciε bilinear n-fold Pfister forms for every i = 1, . . . , r and every ε = (ε1 , . . . , εm ) ∈
{±1}m . By Lemma 35.25,
Xr XX r
ai bi ≡ ai ciε mod I n−1 (F )It (F ).
i=1 ε i=1
(1) (1) (2) (2)
If ε(1) 6= ε(2) in {±1}m then sgnhhε1 d1 , . . . , εm dm ii and sgnhhε1 d1 , . . . , εm dm ii have
disjoint supports on X(F ), hence the same holds for sgn ciε(1) and sgn cjε(2) . It therefore
follows from the hypothesis that
Xr
ai ciε ≡ 0 mod It (F ) for each ε.
i=1
P
Clearly, it suffices to show that ri=1 ai ciε ≡ 0 mod I n−1 (F )It (F ) for each ε.
Fix ε in {±1}m . Suppose that ε 6= (1, . . . , 1). If −1 occurs as the coordinate of ε
corresponding to aji for some j = 1, . . . , r and i = 1, . . . , n then hhε1 d1 , . . . , εm dm ii·bj = 0
in W (F ) and we may assume for all such j that cjε = 0 in W (F ). In particular, if
ε 6= (1, . . . , 1), then
r
X Xr
ai ciε = ai ciε ≡ 0 mod I n−1 (F )It (F )
i=1 i=1
i6=j
by the induction hypothesis. So we may assume that ε = (1, . . . , 1). Then

hhε1 d1 , . . . , εm dm ii ⊗ bi ' hhd1 , . . . , dm ii ⊗ bi ' 2n hhd1 , . . . , dm ii
is independent of i. We therefore may assume that ciε , for i = 1, . . . , r, are all equal to a
single c. Let d = ha1 , . . . , ar i then
r
X
d·c= ai ciε ≡ 0 mod It (F ).
i=1
By Lemma 35.18, we conclude that d · c ∈ I n−1 (F )It (F ) and the theorem follows. ¤
Corollary 35.27. The following are equivalent for a field F of characteristic different
from two:
¡ √ ¢
(1) I n+1 F ( −1) = 0.
(2) F satisfies An+1 and 2I n (F ) = I n+1 (F ).
n n+1
(3) F satisfies An+1 and 2Ired (F ) = Ired (F ).
(4) I (F ) is torsion-free and 2I (F ) = I n+1 (F ).
n+1 n
Proof. (1) ⇒ (2): By√Theorem 35.12, F satisfies An+1 . Theorem 34.22 applied to
the quadratic extension F ( −1)/F gives 2I n (F ) = I n+1 (F ).
n n+1
(2) ⇒ (3) is trivial as 2Ired (F ) = Ired (F ) if 2I n (F ) = I n+1 (F ).
(3) ⇒ (4): As the torsion (n + 1)-fold Pfister forms generate the torsion in I n+1 (F ) by
Theorem 35.26, we have I n+1 (F ) is torsion-free. Suppose that b is an (n + 1)-fold Pfister
form. Then there exist c ∈ I n (F ) and d ∈ Wt (F ) such that b = 2c + d in W (F ). Hence
for some N , we have 2N b = 2N +1 c. As I n+1 (F ) is torsion-free, we have b = 2c in W (F ),
hence bF (√−1) is hyperbolic. By Theorem 34.22, there exists an n-fold Pfister form f such
that b ' 2f. It follows that 2I n (F ) = I n+1 (F ).
√
(4) ⇒ (1) follows from Theorem 34.22 for the quadratic extension F ( −1)/F as forms
in W (K) transfer to torsion forms in W (F ). ¤
¡ √ ¢
Remark 35.28. By Corollary 35.14, the condition that I n+1 F ( −1) = 0 is equiva-
lent to I n+1 (K) = 0 for some quadratic extension K/F . In particular, if I n+1 (K) = 0 for
some quadratic extension K/F then I n+1 (F ) is torsion-free. Much more is true. If K/F
is a field extension of transcendence degree m then I n (K) torsion-free implies n ≥ m and
I n−m (F ) is torsion-free. We shall prove this in Corollary 43.9 below.
Corollary 35.29. Let F be a real closed field and K/F a finitely generated extension
of transcendence degree n. Then I n+1 (K) is torsion-free and 2I n (K) = I n+1 (K).
√ ¡ √ ¢
Proof. As K( −1) is a Cn -field by Theorem 97.7, we have I n+1 K( −1) = 0 and
hence 2I n (K) = I n+1 (K) by Corollary 35.27 applied to the field K. ¤
Corollary 35.30. Let F be a field satisfying I n+1 (F ) = 2I n (F ). Then I n+2 (F ) is
torsion-free.
Proof. If −1 ∈ F 2 then I n+1 (F ) = 0 and the result follows. In particular, we may
assume that char F 6= 2. By Theorem 35.26, it suffices to show that F satisfies An+2 . Let
b be an (n + 2)-fold Pfister form such that 2b = 0 in W (F ). By Lemma 35.2, ¡ we ¢ can
write b = hhwii · c in W (F ) with c an (n + 1)-fold Pfister form and w ∈ D 2h1i . By
assumption, c = 2d in W (F ) for some n-fold Pfister form d. Hence b = 2hhwii · d = 0 in
W (F ). ¤
Remark 35.31. Any local field F satisfies I 3 (F ) = 0 (cf. µ
[90, Cor.¶VI.2.15]). Let
−1, −1
Q2 be the field of 2-adic numbers. Then, up to isomorphism, is the unique
Q2
quaternion algebra (cf. [90, Cor. VI.2.24]) hence I (Q2 ) = 2I(Q2 ) = {0, 4h1i} 6= 0. Thus,
2
in general, I n+2 (F ) cannot be replaced by I n+1 (F ) in the corollary above.
We shall return to these matters in §41.

CHAPTER VI
u-invariants
36. The ū-invariant
Given a field F , it is interesting to see if there exists a uniform bound on the dimension
of anisotropic forms over F , i.e., if there exists an integer n such that every quadratic
form over F of dimension greater than n is isotropic and if such an n exists what is the
minimum. For example, a consequence of the Chevalley-Warning Theorem (cf. [127,
I.2, Th. 3]), is that over a finite field every three dimensional quadratic form is isotropic
and a consequence of the Lang-Nagata Theorem (cf. Theorem 97.7 below) is that every
(2n + 1)-dimensional form over a field of transcendence degree n over an algebraically
closed field is isotropic. When a field is of characteristic different from two, anisotropic
forms are non-degenerate. This is no longer the case for fields of characteristic two. This
leads to the consideration of two invariants. If F is a formally real field then nh1i can
never be isotropic. To obtain meaningful arithmetic data about formally real fields, we
shall strengthen the condition on our forms. Although this makes computation more
delicate, it is a useful generalization. In this section, we shall, for the most part, look at
the simpler case of fields that are not formally real.
Let F be a field. We call a quadratic form ϕ over F locally hyperbolic if ϕFP is
hyperbolic at each real closure FP of F (if any). If F is formally real then the dimension
of every locally hyperbolic form is even. If F is not formally real, every form is locally
hyperbolic. For any field F , a locally hyperbolic form is one that is torsion in the Witt
ring W (F ).
Define the u-invariant of F to be the smallest integer u(F ) ≥ 0 such that every non-
degenerate locally hyperbolic quadratic form over F of dimension > u(F ) is isotropic (or
infinity if no such integer exists) and the ū-invariant of F to be the smallest integer
ū(F ) ≥ 0 such that every locally hyperbolic quadratic form over F of dimension > ū(F )
is isotropic (or infinity if no such integer exists). The u-invariant above was first defined
in [41] and the ū-invariant by Baeza in [16].
Remark 36.1. (1) We have ū(F ) ≥ u(F ).
(2) If char F 6= 2, every anisotropic form is non-degenerate hence ū(F ) = u(F ).
(3) If F is formally real, the integer ū(F ) = u(F ) is even.
(4) As any (non-degenerate) quadratic form contains (non-degenerate) subforms of all
smaller dimensions, if F is not formally real, we have u(F ) ≤ n if and only if every non-
degenerate quadratic form of dimension n + 1 is isotropic and ū(F ) ≤ n if and only if
every quadratic form of dimension n + 1 is isotropic.
163
164 VI. u-INVARIANTS
Example 36.2. (1) If F is a formally real field then ū(F ) = 0 if and only if F is
pythagorean.
(2) Suppose that F is a quadratically closed field. If char F 6= 2 then ū(F ) = 1 as every
form is diagonalizable. If char F = 2 then ū(F ) ≤ 2 with equality if F is not separably
closed by Example 7.34.
(3) If F is a finite field then ū(F ) = 2.
(4) Let F be a field of characteristic not two and ρ a quadratic form over the Laurent
series field F ((t)). As ρ is diagonalizable, ρ ∼ = ϕ ⊥ tψ for some quadratic forms ϕ and ψ
over F . Using Lemma 19.5, we deduce that ρ is anisotropic if and only if both ϕ and ψ
are. Moreover, one can check that ¡ ρ is¢ torsion if and only if both ϕ and ψ are torsion. If
ū(F ) is finite, it follows that ū F ((t)) = 2ū(F ). This equality is also true if char F = 2
(cf. [16]).
(5) If F is a Cn -field then ū(F ) ≤ 2n .
(6) If F is a local field then ū(F ) = 4. If char F > 0 then u(F ) = 4 by Example (4). If
char F = 0 this follows as in the argument for Example (4) using also Lemma 19.4 (cf.
also [115]).
(7) If F is a global field then ū(F ) = 4. If char F = 0 this follows from the Hasse-
Minkowski Theorem [90], VI.3.1. If char F > 0 then F is a C2 -field by Theorem 97.7.
Proposition 36.3. Let F be a field with Iq3 (F ) = 0. If 1 < u(F ) < ∞ then u(F ) is
even.
Proof. The hypothesis implies that F is not formally real. Suppose that u(F ) > 1
is odd and let ϕ be a non-degenerate anisotropic quadratic form with dim ϕ = u(F ).
We claim that ϕ ' ψ ⊥ h−ai for some ψ ∈ Iq2 (F ) and a ∈ F × . If char F 6= 2 then
ϕ ⊥ hai ∈ Iq2 (F ) for some a ∈ F × by Proposition 4.13. This form is isotropic, hence
ϕ ⊥ hai ' ψ ⊥ H for some ψ ∈ Iq2 (F ) and therefore ϕ ' ψ ⊥ h−ai. If char F = 2 write
ϕ ' µ ⊥ hai for some form µ and a ∈ F × . Choose b ∈ F such that the discriminant of
the form µ ⊥ [a, b] is trivial, i.e., µ ⊥ [a, b] ∈ Iq2 (F ). By assumption the form µ ⊥ [a, b] is
isotropic, i.e., µ ⊥ [a, b] ' ψ ⊥ H for a form ψ ∈ Iq2 (F ). It follows from (8.7) that
ϕ ' µ ⊥ hai ∼ µ ⊥ [a, b] ⊥ hai ∼ ψ ⊥ hai,
hence ϕ ' ψ ⊥ hai as these forms have the same dimension. This proves the claim.
Let b ∈ D(ψ). As hhabii ⊗ ψ ∈ Iq3 (F ) = 0, we have ab ∈ G(ψ). Therefore, a = ab/b ∈
D(ψ) and hence the form ϕ is isotropic, a contradiction. ¤
Corollary 36.4. The u-invariant of a field is not equal to 3, 5 or 7.
Let r > 0 be an integer. Define the ūr -invariant of F to be the smallest integer
ūr (F ) ≥ 0 such that every set of r quadratic forms on a vector space over F of dimension
> ūr (F ) has common nontrivial zero.
In particular, if ūr (F ) is finite then F is not a formally real field. We also have
ū1 (F ) = ū(F ) when F is not formally real. The following is due to Leep (cf. [94]):
Theorem 36.5. Let F be a field. Then for every r > 1, we have
ūr (F ) ≤ rū1 (F ) + ūr−1 (F ).
36. THE ū-INVARIANT 165
Proof. We may assume that ūr−1 (F ) is finite. Let ϕ1 , . . . , ϕr be quadratic forms on

a vector space V over F of dimension n > rū1 (F ) + ūr−1 (F ). We shall show that the
forms have an isotropic vector in V . Let W be a totally isotropic subspace of V of the
forms ϕ1 , . . . , ϕr−1 of the largest dimension d. Let Vi be the orthogonal complement of
W in V relative to ϕi for each i = 1, . . . , r − 1. We have dim Vi ≥ n − d.
Let U = V1 ∩ · · · ∩ Vr−1 . Then W ⊂ U and dim U ≥ n − (r − 1)d. Choose a subspace
U 0 ⊂ U such that U = W ⊕ U 0 . We have
¡ ¢
dim U 0 ≥ n − rd > r ū1 (F ) − d + ūr−1 (F ).
If d ≤ ū1 (F ) then dim U 0 > ūr−1 (F ), hence the forms ϕ1 , . . . , ϕr−1 have an isotropic
vector u ∈ U 0 . Then the subspace W ⊕F u is totally isotropic for these forms, contradicting
the maximality of W .
It follows that d > ū1 (F ). The form ϕr therefore has an isotropic vector in U 0 which
is isotropic for all the ϕi ’s. ¤
Corollary 36.6. If F is not formally real then ūr (F ) ≤ 12 r(r + 1)ū(F ).
This improves the linear bound in [34] to:
Corollary 36.7. Let K/F be a finite field extension of degree r. If F is not formally
real then ū(K) ≤ 12 (r + 1)ū(F ).
Proof. Let s1 , s2 , . . . , sr be a basis for the space of F -linear functionals on K. Let
ϕ be a quadratic form over K of dimension n > 12 (r + 1)ū(F ). As dim(si )∗ (ϕ) = rn >
1
2
r(r + 1)ū(F ) for each i = 1, . . . , r, by Corollary 36.6, the forms (si )∗ (ϕ) have a common
isotropic vector which is then an isotropic vector for ϕ. ¤
Let K/F be a finite extension with F not formally real. We shall show that if ū(K) is
finite then so is ū(F ). We begin with the case that F is a field of characteristic two with
an observation by Mammone, Moresi, and Wadsworth (cf. [97]).
Lemma 36.8. Let F be a field of characteristic two. Let ϕ be an even dimensional
non-degenerate quadratic form over F and ψ a totally singular quadratic form over F . If
ϕ ⊥ ψ is anisotropic then
1
dim ϕ + dim ψ ≤ [F : F 2 ].
2
Proof. Let ϕ ' [a1 , b1 ] ⊥ · · · ⊥ [am , bm¡] with¢ai , bi ∈ F and ψ ' hc1 , . . . , cn i with
ci ∈ F × . For each i = 1, . . . , m let di ∈ D [ai , bi ] . Then {c1 , . . . , cn , d1 , . . . dm } is F 2 -
linearly independent. The result follows. ¤
Proposition 36.9. Let F be a field of characteristic two and K/F a finite extension.
Then
ū(F ) ≤ 2ū(K) ≤ 4ū(F ).
Proof. If c1 , . . . , cn are F 2 -linearly independent then hc1 , . . . , cn i is anisotropic. By
the lemma, it follows that we have
[F : F 2 ] ≤ ū(F ) ≤ 2[F : F 2 ].
As [F : F 2 ] = [K : K 2 ] (cf. (35.6)), we have

ū(F ) ≤ 2[F : F 2 ] = 2[K : K 2 ] ≤ 2ū(K) ≤ 4[K : K 2 ] = 4[F : F 2 ] ≤ 4ū(F ). ¤
Remark 36.10. Let F be a field of characteristic two. The proof above shows that
every anisotropic totally singular quadratic form has dimension at most [F : F 2 ] and
if [F : F 2 ] is finite then there exists an anisotropic totally singular quadratic form of
dimension [F : F 2 ].
Remark 36.11. Let F be a field of characteristic two such that [F : F 2 ] is infinite
but F separably closed. Then ū(F ) is infinite but u(F ) = 1 by Exercise 7.35.
We now look at finiteness of ū coming down from a quadratic extension following the
ideas in [41].
Proposition 36.12. Let K/F be a quadratic extension with F not formally real. If
ū(K) is finite then ū(F ) < 4ū(K).
Proof. If char F = 2 then ū(F ) ≤ 2ū(K) by Proposition 36.9, so we may assume
that char F 6= 2. We first show that ū(F ) is finite. Let ϕ be an anisotropic quadratic form
over F . By Proposition 34.8, there exist quadratic forms ϕ1 and µ0 over F with (µ0 )K
anisotropic satisfying
ϕ ' hhaii ⊗ ϕ1 ⊥ µ0
×
√
where a ∈ F satisfies K = F ( a). In particular, dim(µ0 ) ≤ ū(K). Analogously, there
exist quadratic forms ϕ2 and µ1 over F with (µ1 )K anisotropic satisfying
ϕ1 ' hhaii ⊗ ϕ2 ⊥ µ1 .
Hence
¡ ¢
ϕ ' hhaii ⊗ hhaii ⊗ ϕ2 ⊥ µ1 ⊥ µ0 ' 2hhaii ⊗ ϕ2 ⊥ hhaii ⊗ µ1 ⊥ µ0
as hha, aii = 2hhaii. Continuing in this way, we see that
ϕ ' 2k−1 hhaii ⊗ ϕk ⊥ 2k−2 hhaii ⊗ µk−1 ⊥ · · · ⊥ hhaii ⊗ µ1 ⊥ µ0
for some forms ϕk and µi over F satisfying dim µi ≤ ū(K) for all i. By Proposition 31.4,
there exists an integer n such that 2n hhaii = 0 in W (F ). It follows that
dim ϕ ≤ (2n + · · · + 2 + 1)ū(K) ≤ 2n+1 ū(K)
hence dim ϕ is finite.
We now show that ū(F ) < 4ū(K). As ū(F ) is finite, there exists an anisotropic form ϕ
over F of dimension ū(F ). Let s : K → F be a non-trivial F -linear functional satisfying
s(1) = 0. We can write
ϕ ' µ ⊥ s∗ (ψ)
with quadratic forms ψ over K and µ over F satisfying µ ⊗ NK/F is anisotropic by
Proposition 34.6. Then
1
dim s∗ (ψ) ≤ 2ū(K) and dim µ ≤ ū(F ).
2
If
¡ dim s (ψ)
∗ ¢ = 2ū(K) then ψ is a ū(K)-dimensional form over K hence universal as every
ū(K) + 1 -dimensional form is isotropic over the non formally real field K. In particular,
37. THE u-INVARIANT FOR FORMALLY REAL FIELDS 167
ψ ' hxiK ⊥ ψ1 for some x ∈ F × . Thus s∗ (ψ) = s∗ (ψ1 ) in W (F ) so s∗ (ψ) is isotropic, a

contradiction. Therefore, we have dim s∗ (ψ) < 2ū(K), hence
2ū(K) > dim s∗ (ψ) = dim ϕ − dim µ ≥ ū(F ) − ū(F )/2 ≥ ū(F )/2.
The result follows. ¤
Proposition 36.13. Let K/F be a finite extension with F not formally real. Then
ū(F ) is finite if and only if ū(K) is finite.
Proof. If char F = 2, the result follows by Proposition 36.9, so we may assume that
char F 6= 2. By Corollary 36.7, we need only show if ū(K) is finite then ū(F ) is also
finite. Let L be the normal closure of K/F and E0 the fixed field of the Galois group of
L/F . Then E0 /F is of odd degree as char F 6= 2. Let E be the fixed field of a Sylow
2-subgroup of the Galois group of L/F . Then E/F is also of odd degree. Therefore, if
ū(E) is finite so is ū(F ) by Springer’s Theorem 18.5. Hence we may assume that E = F .
By Corollary 36.7, we have ū(L) is finite so we may assume that L = K and K/F is a
Galois 2-extension. By induction on [K : F ], we may assume that K/F is a quadratic
extension, the case established in Proposition 36.12. ¤
Let K/F be a normal extension of degree 2m r with r odd and F not formally real.
If ū(K) is finite the argument in Proposition 36.13 and the bound in Proposition 36.12
shows that ū(F ) ≤ 4r ū(K). We shall improve this bound in Remark 37.6 below.
37. The u-invariant for formally real fields
If F is formally real and K/F finite then ū(K) can be infinite and ū(F ) finite. Indeed,
let F0 be the euclidean field of real constructible numbers. Then there exist extensions
Er /F0 of degree r none of which are both pythagorean and formally real. In partic-
ular, ū(Er ) > 0. It is easy to see that ū(Er ) ≤ 4. (In fact, it can be shown that
ū(Er ) ≤ 2.) For example, E2 is the quadratic closure of the rational numbers. Let
F = F0 ((t1 )) · · · ((tn )) · · · denote the field of iterated Laurent series in infinitely many
variables over F0 . Then F is pythagorean by Example 36.2(1) so ū(F ) = 0. However,
Kr = Er ((t1 )) · · · ((tn )) · · · has infinite u-invariant by Example 36.2(4). In fact, in [44]
for each positive integer n, formally real fields Fn are constructed with ū(Fn ) = 2n and
having a formally real quadratic extension K/Fn with ū(K) = ∞ and formally fields Fn0
are constructed with ū(Fn0 ) = 2n and such that every finite non-formally real extension L
of F has infinite ū-invariant.
However, we can determine when finiteness of the ū-invariant persists when going up
a quadratic extension and when coming down one. Since we already know this when the
base field is not formally real, we shall mostly be interested in the formally real case. In
particular, we shall assume that the fields in this section are of characteristic different
from two and hence the ū-invariant and u-invariant are identical.
We need some preliminaries.
√
Lemma 37.1. Let F be a field of characteristic different from
¡ two
¢ and K = F ( a) a
quadratic extension of F . Let b ∈ F × \ F ×2 and ϕ ∈ annW (F ) hhbii be anisotropic. Then
ϕ = ϕ1 + ϕ2 in W (F ) for some forms ϕ1 and ϕ2 over F satisfying
¡ ¢
(1) ϕ1 ∈ hhaiiW (F¡ ) ∩ ann
¢ W (F ) hhbii is anisotropic.
(2) ϕ2 ∈ annW (F ) hhbii .
(3) (ϕ2 )K is anisotropic.
Proof. By Corollary 6.23 the dimension of ϕ is even. We induct on dim ϕ. If
ϕK is hyperbolic then ϕ = ϕ1 works by Corollary 34.12 and if ϕK is anisotropic then
ϕ = ϕ2 works. So we may assume that ϕK is isotropic but not hyperbolic. In particular,
dim ϕ ≥ 4. By Proposition 34.8, we can write
ϕ ' xhhaii ⊥ µ
×
¡ ¢
for some x ∈ F and even dimensional form form ¡ µ over ¢F . As ϕ ∈ ann W (F ) hhbii ,
we have hhbii · µ = −xhhb, aii in W (F ), so dim hhbii ⊗ µ an = 0 or 4. Therefore, by
Proposition 6.25, we can write
µ ' µ1 ⊥ yhhcii
¡ ¢
for some y, c ∈ F × and even dimensional form µ1 ∈ annW (F ) hhbii¡ . Substituting
¢ in the
previous isometry and taking determinants,
¡ ¢ we see that ac ∈ D hhbii by Proposition
6.25. Thus c = az for some z ∈ D hhbii . Consequently,
ϕ ' xhhaii ⊥ yhhazii ⊥ µ1 = xhha, −xyzii + yhhzii + µ1
¡ ¢ ¡ ¢
in W (F ). Let µ2 ' yhhzii ⊥ µ1 an . As yhhzii lies in annW (F ) hhbii , so does µ2 and
hence also xhha, −xyzii. By induction on dim ϕ, we can write µ2 = ϕ e1 + ϕe2 in W (F )
where ϕ
e1 satisfies condition (1) and ϕ
e2 satisfies conditions (2) and (3). It follows that
¡ ¢
ϕ1 ' hha, −xyzii ⊥ ϕ e1 an and ϕ2 ' ϕ e2
work. ¤
Exercise 37.2. Let ϕ and ψ be 2-fold Pfister forms over a field of characteristic not
2. Prove that the group ϕW (F ) ∩ annW (F ) (ψ) ∩ I 2 (F ) is generated by 2-fold Pfister forms
ρ in annW (F ) (ψ) that are divisible by ϕ. This exercise generalizes. (Cf. Exercise 41.8
below.)
As we are interested in the case of fields of characteristic not¡ two, ¢ we shall show
to test finiteness of the u-invariant, it suffices to look at annW (F ) 2h1i . For a field of
characteristic not two, let
u0 (F ) := max{dim ϕ | ϕ is an anisotropic form over F and 2ϕ = 0 in W (F )}
or ∞ if no such maximum exists.
Lemma 37.3. Let F be a field of characteristic not two. Then u0 (F ) is finite if and
only if u(F ) is finite. Moreover, if u(F ) is finite then u(F ) = u0 (F ) = 0 or u0 (F ) ≤
u(F ) < 2u0 (F ).
Proof. We may assume that u0 (F ) > 0, i.e., that F is not a formally real pythagorean
0
field. Let ϕ be an n-dimensional anisotropic form over F . Suppose
¡ ¢that n ≥ 2u (F ). By
Proposition 6.25, we can write ϕ ' µ1 ⊥ ϕ1 with µ1 ∈ annW (F ) 2h1i and 2ϕ1 anisotropic.
By assumption, dim µ1 ≤ u0 (F ). Thus
2u0 (F ) ≤ dim ϕ = dim µ1 + dim ϕ1 ≤ u0 (F ) + dim ϕ1
hence 2u0 (F ) ≤ dim 2ϕ1 . As (2ϕ)an ' 2ϕ1 , we have dim(2ϕ)an ≥ 2u0 (F ). Repeating the
argument, we see inductively that dim(2m ϕ)an ≥ 2u0 (F ) for all m. In particular, ϕ is not
torsion. The result follows. ¤
Hoffmann has shown that there exist fields F satisfying u0 (F ) < u(F ). (Cf. [55].)
Let K/F be a quadratic extension. As it is not true that u(F ) is finite if and only
if u(K) is when F is formally real, we need a further condition for this to be true. This
condition is given by a relative u-invariant.
Let F be a field of characteristic not two and L/F a field extension. The relative
u-invariant of L/F is defined as
u(L/F ) := max{dim(ϕL )an | ϕ a quadratic form over F with ϕL torsion in W (L)}
or ∞ if no such integer exists.
The introduction of this invariant and its study come from Lee in [92] and [32]. We
shall prove
Theorem 37.4. Let F be a field of characteristic different than two and K a quadratic
extension of F . Then u(F ) and u(K/F ) are both finite if and only if u(K) is finite.
Moreover, we have:
(1) If u(F ) and u(K/F ) are both finite then u(K) ≤ u(F ) + u(K/F ). If, in addition,
K is not formally real then u(K) ≤ 21 u(F ) + u(K/F ).
(2) If u(K) is finite then u(K/F ) ≤ u(K) and u(F ) < 6u(K) or u(F ) = u(K) = 0.
If, in addition, K is not formally real then u(F ) < 4u(K).
√
Proof. Let K = F ( a) and s∗ : W (K) √→ W (F ) be the transfer induced by the
F -linear functional defined by s(1) = 0 and s( a) = 1.
Claim: Let ϕ be an anisotropic quadratic form over K such that s∗ (ϕ) is torsion in W (F ).
Then there exist a form σ over F and a form ψ over K satisfying
(a) dim σ = dim ϕ.
(b) ψ is a torsion form in W (K).
(c) dim ψ ≤ 2 dim ϕ and ϕ ' (σK ⊥ ψ)an .
(d) If s∗ (ϕ) is anisotropic over F then dim ϕ ≤ dim ψ.
In particular, if u(F ) is finite and s∗ (ϕ) anisotropic and torsion then dim ϕ ≤ 21 u(F ) and
dim ψ ≤ u(F ).
Let 2n s∗ (ϕ) = 0 in W (F ) for some integer n. By Corollary 34.3 with ρ = 2n h1i, there
exists a form σ over F such that dim σ = dim ϕ and 2n ϕ ' 2n σK . Let ψ ' (ϕ ⊥ (−σ))an .
Then ψ is a torsion form in W (K) as it has trivial total signature. Condition (c) holds
by construction and (d) holds as s∗ (ψ) = s∗ (ϕ) in W (F ).
We now prove (1). Suppose that both u(F ) and u(K/F ) are finite. Let τ be an
anisotropic torsion form over K. By Proposition 34.1, there exists an isometry τ ' ϕ ⊥ µK
for some form τ over K satisfying s∗ (ϕ) is anisotropic and some form µ over F . As
s∗ (ϕ) = s∗ (τ ) is torsion, we can apply the claim to ϕ. Let σ over F and ψ over K
be forms as in the claim. By the last statement of the claim, we have dim ϕ ≤ 21 u(F ).
In particular, we have dim ψ ≤ 2 dim ϕ ≤ u(F ) and ϕ = ψ + σK in W (K). Since τ
¡ ¢
and ψ are torsion so is (σ + µ)K . As τ = ψ + (σ ⊥ µ)K an in W (K), it follows that
dim τ ≤ u(F ) + u(K/F ) as needed.
Finally, if K is not formally real then as above, we have τ ' ϕ ⊥ µK with dim ϕ ≤
1
2
u(F ). As every F -form is torsion in W (K), we have dim µK ≤ u(K/F ) and the proof
of (1) is complete.
We now prove (2). Suppose that u(K) is finite. Certainly u(K/F ) ≤ u(K). We show
the rest of the first
¡ statement.
¢ By Lemma 37.3, it suffices to show that u0 (F ) ≤ 3u0 (K).
Let ϕ ∈ annW (F ) 2h1i be anisotropic. By Lemma ¡ 37.1
¢ and Corollary 34.33, we can
decompose ϕ = ϕ1 + ϕ2 in W (F ) with ϕ2 ∈ annW (F ) 2h1i satisfying (ϕ2 )K is anisotropic
and ϕ1 is anisotropic over F and lies in
¡ ¢ ¡ ¢ ¡ ¢
hhaiiW (F ) ∩ annW (F ) 2h1i ⊂ annW (F ) hhaii ∩ annW (F ) 2h1i
¡ ¢
using Lemma 34.33. In particular, (ϕ2 )K ∈ annW (K) 2h1i so dim ϕ2 ≤ u0 (K). Conse-
quently, to show that u0 (F ) ≤ 3u0 (K), it suffices to show dim ϕ1 ≤ 2u0 (K). This follows
from (i) of the following (with σ = ϕ1 ):
Claim: Let σ be a non-degenerate quadratic form over F .
¡ ¢ ¡ ¢
(i) If σ ∈ annW (F ) hhaii ∩ annW (F ) 2h1i then dim σan ≤ 2u0 (K).
¡ ¢
(ii) If σ ∈ annW (F ) hhaii ∩ Wt (F ) then dim σan ≤ 2u(K) with inequality if K is not
formally real.
¡ ¢
By Corollary 34.33, in the situation of (i), there exists τ ∈ annW (K) 2h1i such that
σ = s∗ (τ ). Then dim σan ≤ dim s∗ (τan ) ≤ 2 dim τan ≤ 2u0 (K) as needed.
We turn to the proof of (ii) which implies the the bound on u(F ) in statement (2) for
arbitrary K (with σ = ϕ1 ). In the situation of (ii), we have dim σan ≤ u(K) by Corollaries
34.12 and 34.32. If K is not formally real then any u(K)-dimensional form τ over K is
universal. In particular, D(τ ) ∩ F × 6= ∅ and (ii) follows.
Now assume that K is not formally real. ¡ Let¢ ϕ be an anisotropic torsion form over
F of dimension u(F ). As im(s∗ ) = annW (F ) hhaii by Corollary 34.12, using Proposition
6.25, we have a decomposition ϕ ' ϕ3 ⊥ ϕ4 with ϕ4 a form over F satisfying hhaii ⊗ ϕ4
is anisotropic and ϕ3 ' s∗ (τ ) for some form τ over K. Since ϕ3 lies in
¡ ¢ ¡ ¢ ¡ ¢
s∗ W (K) = s∗ Wt (K) = annW (F ) hhaii ∩ Wt (F )
by Corollaries 34.12 and 34.32, we have dim ϕ3 < 2u(K) by the claim above. As hhaii·ϕ4 =
hhaii·ϕ in W (F ) hence is torsion, we have dim ϕ4 ≤ u(F )/2. Therefore, 2u(K) > dim ϕ3 =
dim ϕ − dim ϕ4 ≥ u(F ) − u(F )/2 and u(F ) < 4u(K). ¤
√
Of course, by Theorem 36.6 if F is not formally real and K = F ( a) is a quadratic
extension then u(K) ≤ 32 u(F ).
Corollary 37.5. Let F be a field of transcendence degree n over a real closed field.
Then u(F ) < 2n+2 .
√
Proof. F ( −1) is a Cn -field by Corollary 97.7. ¤
Remark 37.6. Let F be a field of characteristic not two and K/F a finite normal
extension. Suppose that u(K) is finite. If K/F is quadratic then the proof of Theorem
37.4 shows that u0 (F ) ≤ 3u0 (K). If K/F is of degree 2r m with m odd, arguing as in
Proposition 36.13, shows that u0 (F ) ≤ 3r u0 (K) hence u(F ) ≤ 2 · 3r u(K).
One case where the bound in the remark can be sharpened is the following which
generalizes the case of a pythagorean field of characteristic different from two.
Proposition 37.7. Let F be a field of characteristic different from two and K/F a
finite normal extension. If u(K) ≤ 2 then u(F ) ≤ 2.
Proof. By Proposition 35.1, we know for a field E that I 2 (E) is torsion-free if and
only if E satisfies A2 , i.e., there are no anisotropic 2-fold torsion Pfister forms. In par-
ticular, as u(K) ≤ 2, we have I 2 (K)√ is torsion-free. Arguing as in Proposition 36.13, we
reduce to the case that K = F ( a) is a quadratic extension of F , hence I 2 (F ) is also
torsion-free
¡ by ¢ Theorem 35.12. It follows that every torsion element ρ in I(F ) lies in
2
annW (F ) 2h1i . ¡In particular,
¢ by Proposition 6.25,
¡ we can
¢ write ρ ' hhwii mod I (F )
for some w ∈ D 2h1i hence ρ ' hhwii some w ∈ D 2h1i and is universal. Consequently,
every even dimensional anisotropic torsion form over F is of dimension at most two. Sup-
pose that there exists an odd dimensional anisotropic torsion form ϕ over F . Then F is
not formally real hence all forms are torsion. As every two dimensional form over F is
universal by the above, we must have dim ϕ = 1. The result follows. ¤
Corollary 37.8. Let F be a field of transcendence degree one over a real closed field.
Then u(F ) ≤ 2.
×
Exercise
√ 37.9. Let F be a field of arbitrary characteristic and a ∈ F totally positive.
If K = F ( a) then u(K) ≤ 2u(F ).
We next show if K/F is a quadratic extension with K not formally real then the
relative u-invariant already determines finiteness. We note
√
of F that is not formally real. If ϕ is a non-degenerate quadratic form over F then,
by Proposition 34.8, there exist forms ϕ1 and ψ such that ϕ ' hhaii ⊗ ψ ⊥ ϕ1 with
dim ϕ1 ≤ u(K/F ).
We need the following simple lemma.
√
Lemma 37.11. Let F be a field of characteristic different from two and K = F ( a)
a quadratic extension of F that is not formally real. Suppose that u(K/F ) < 2m . Then
I m+1 (F ) is torsion-free, I m+1 (K) = 0, and the exponent of Wt (F ) is at most 2m+1 .
∗
Proof. If ρ ∈ Pm (F ) then rK/F (ρ) = 0 as K is not formally real. So I m (F ) =
hhaiiI¡ m−1√(F ) by
¢ Theorem 34.22. It follows that I m+1 (K) = 0 by Lemma 34.16. Hence
I m+1 F ( −1) = 0 by Corollary 35.14. The result follows by Corollary 35.27. ¤
If F is a local field in the above then one can show that u(K/F ) = 2 for any quadratic
extension K of F but neither I 2 (F ) nor I 2 (K) is torsion-free.
Theorem 37.12. Let F be a field of characteristic different from two. Suppose that
K is a quadratic extension of F with K not formally real. Then u(K/F ) is finite if and
only if u(K) is finite.
Proof. By Theorem 37.4, we may assume that u(K/F ) is finite and must show that
u(F ) is also finite. Let ϕ be an anisotropic form over F satisfying 2ϕ = 0 in W (F ).
By the lemma, I n−1 (F ) is torsion-free for some n ≥ 1. We apply the Remark 37.10
iteratively. In particular, if dim ϕ is large then ϕ ' xρ ⊥ ψ for some ρ ∈ Pn (F ) (cf.
the proof of Proposition 36.12). Indeed, computation shows that if u(K/F ) < 2m and
dim ϕ > 2m (2m+2 − 1) then n = m + 2 works. As ρ is an anisotropic Pfister form and
I n (F ) is torsion-free, 2ρ is also anisotropic. Scaling ϕ, we may assume that x = 1. Write
ψ ' ϕ1 ⊥ ϕ2 with 2ϕ1 = 0 in W (F ) and 2ϕ2 anisotropic. Then we have 2ρ ' 2(−ϕ2 ).
If b ∈ D(−ϕ2 ) then 2hhbii · ρ is isotropic hence is zero in W (F ). As I n (F ) is torsion-
free, hhbii · ρ = 0 in W (F ) and b ∈ D(ρ). It follows that ϕ cannot be anisotropic if
dim ϕ > 2m (2m+2 − 1). By Lemma 37.3, it follows that u(F ) ≤ 2m+1 (2m+2 − 1) and the
result follows by Theorem 37.4. ¤
The bounds in the proof can be improved but are still very weak. The theorem does
not generalize to the case when K is formally real. Indeed let F0 be a formally real subfield
of the algebraic closure of the rationals having square classes represented√by ±1, ±w where
w is a sum of (two) squares. Let F = F0 ((t1 ))((t2 )) · · · and K = F ( w). Then, using
Corollary 34.12, we see that u(K/F ) = 0 but both u(F ) and u(K) are infinite.
Corollary 37.13. Let F be a field of characteristic ¡ different
√ ¢ from two. Then u(K)
is finite for all finite extensions of F if and only if u F ( −1) is finite if and only if
√
u(F ( −1)/F ) is finite.
Remark 37.14. Using the theory of abstract Witt rings, Schubert proved that if F is
a formally real field then u(K) is finite for all finite extensions
¡ √of F ¢if and only if u(F ) is
n+1 n
finite and Ired (F ) = 2Ired (F ) for some n if and only if u F ( −1) is finite (cf. [126]).
Recall that the condition on the reduced Witt ring is equivalent to the cokernel of the
total signature map having finite exponent by Proposition 35.23.
38. Construction of fields with even u-invariant
In 1953 Kaplansky conjectured in [69] that u(F ) if finite was always a power of two.
This was shown to be false in 1989 in [103] where it was shown there exist fields having
u-invariant six and afterwards in [104] having u-invariant any even integer. Subsequently,
Izhboldin constructed fields having u-invariant 9 in [66] and Vishik has constructed fields
having u-invariant 2r + 1 for every r ≥ 3 in [141].
By taking iterated Laurent series fields over the complex numbers, we can construct
fields whose u-invariant is 2n for any n ≥ 0. (We also know that formally real pythagorean
fields have u-invariant zero.) In this section, given any even integer m > 0, we construct
fields whose u-invariant is m. This was construction was first done in [104] with the
simplified proof here due to Tignol (cf. [136]).
Lemma 38.1. Let ϕ ∈ Iq2 (F ) be a form of dimension 2n ≥ 2. Then ϕ is a sum of
n − 1 general quadratic 2-fold Pfister forms in Iq2 (F ) and ind clif(ϕ) ≤ 2n−1 .
38. CONSTRUCTION OF FIELDS WITH EVEN u-INVARIANT 173
Proof. We induct on n. If n = 1, we have ϕ = 0 and the statement is clear. If n = 2,

ϕ is a general 2-fold Pfister form and by Proposition 12.4, we have clif(ϕ) = [Q], where
Q is a quaternion algebra such that NrdQ is similar to ϕ. Hence ind clif(ϕ) ≤ 2.
In the case n ≥ 3 write ϕ = σ ⊥ ψ where σ is a binary form. Choose a ∈ F × such
that the form aσ ⊥ ψ is isotropic, i.e., aσ ⊥ ψ ' H ⊥ µ for some form µ of dimension
2n − 2. We have in Iq (F ):
ϕ = σ + ψ = hhaiiσ + µ
¡ ¢
and therefore clif(ϕ) = clif hhaiiσ · clif(µ) by Lemma 14.2. Applying the induction
hypothesis to µ, we have ϕ is a sum of n − 1 general quadratic 2-fold Pfister forms and
¡ ¢
ind clif(ϕ) ≤ ind clif hhaiiσ · ind clif(µ) ≤ 2 · 2n−2 = 2n−1 . ¤
Corollary 38.2. In the condition of the lemma assume that ind clif(ϕ) = 2n−1 . Then
ϕ is anisotropic.
Proof. Suppose ϕ is isotropic, i.e., ϕ ' H ⊥ ψ for some ψ of dimension 2n − 2.
Applying Lemma 38.1 to ψ, we have ind clif(ϕ) = ind clif(ψ) ≤ 2n−2 , a contradiction. ¤
Lemma 38.3. Let D be a tensor product of n − 1 quaternion algebras (n ≥ 1). Then
there is a ϕ ∈ Iq2 (F ) of dimension 2n such that clif(ϕ) = [D] in Br(F ).
Proof. We induct on n. The case n = 1 follows from Proposition 12.4. If n ≥ 2
write D = Q ⊗ B, where Q is a quaternion algebra and B is a tensor product of n − 2
quaternion algebras. By the induction hypothesis, there is ψ ∈ Iq2 (F ) of dimension 2n − 2
such that clif(ψ) = [B]. Choose a quadratic 2-fold Pfister form σ with clif(σ) = [Q] and
an element a ∈ F × such that aσ ⊥ ψ is isotropic, i.e., aσ ⊥ ψ ' H ⊥ ϕ for some ϕ of
dimension 2n. Then ϕ works as clif(ϕ) = clif(σ) · clif(ψ) = [Q] · [B] = [D]. ¤
Let A be a set (of isometry classes) of irreducible quadratic forms. For any finite
subset S ⊂ A let XS be the product of all the quadrics Xϕ with ϕ ∈ S. If S ⊂ T are
two subsets of A we have the dominant projection XT → XS and therefore the inclusion
of function fields F (XS ) → F (XT ). Set FA = colim FS over all finite subsets S ⊂ A. By
construction, all quadratic forms ϕ ∈ A are isotropic over the field extension FA /F . A
field is called 2-special if every finite extension of it has degree a power of 2 (cf. §101.B).
Theorem 38.4. Let F be a field and n ≥ 1 an integer. Then there is a field extension
E of F satisfying
(1) u(E) = 2n.
(2) Iq3 (E) = 0.
(3) E is 2-special.
Proof. To every field L, we associate three fields L(1) , L(2) , and L(3) as follows:
Let A be the set (of isometry classes) of all non-degenerate quadratic forms over L
of dimension 2n + 1. We set L(1) = LA . Every non-degenerate quadratic form over L of
dimension 2n + 1 is isotropic over L(1) .
Let B be the set (of isometry classes) of all quadratic 3-fold Pfister forms over L. We
set L(2) = LB . By construction, every quadratic 3-fold Pfister form over L is isotropic
over L(2) .
Finally let L(3) be a 2-special closure of L (cf. §101.B).

Let D be a central division L-algebra of degree 2n−1 . By Corollaries 30.9, 30.11, and
98.6, the algebra D remains a division algebra when extended to L(1) , L(2) , or L(3) .
Let L be a field extension of F such that there is a central division algebra D over L
that is a tensor product of n − 1 quaternion algebras (cf. Proposition 98.19). By Lemma
38.3, there is a ϕ ∈ Iq2 (L) of dimension 2n such that clif(ϕ) = [D] in Br(L).
We construct a tower of field extensions E0 ⊂ E1 ⊂ E2 ⊂ . . . by induction. We set
E0 = L. If Ei is defined we set Ei+1 = (((Ei )(1) )(2) )(3) . Note that the field Ei+1 is 2-special
and all non-degenerate quadratic forms of dimension 2n + 1 and all 3-fold Pfister forms
over Ei are isotropic over Ei+1 . Moreover the algebra D remains a division algebra over
Ei+1 .
Now set E = ∪Ei . Clearly E has the following properties:
(i) All (2n + 1)-dimensional Pfister forms over E are isotropic. In particular, u(E) ≤
2n.
(ii) The field E is 2-special.
(iii) All quadratic 3-fold Pfister forms over E are isotropic. In particular Iq3 (E) = 0 .
(iv) The algebra DE is a division algebra.
As clif(ϕE ) = [DE ], it follows from Corollary 38.2 that ϕE is anisotropic. In particular,
u(E) = 2n. ¤
39. Addendum: Linked fields and the Hasse number
Theorem 39.1. Let F be a field. Then the following conditions are equivalent:
(1) Every pair of quadratic 2-fold Pfister forms over F are linked.
(2) Every 6-dimensional form in Iq2 (F ) is isotropic.
(3) The tensor product of two quaternion algebras over F is not a division algebra.
(4) Every two division quaternion algebras over F have isomorphic separable qua-
dratic subfields.
(5) Every two division quaternion algebras over F have isomorphic quadratic sub-
fields.
(6) The classes of quaternion algebras in Br(F ) form a subgroup.
Proof. (1) ⇒ (2): Let ψ be a 6-dimensional form in Iq2 (F ). By Lemma 38.1, we have
ψ = ϕ1 + ϕ2 , where ϕ1 and ϕ2 are general quadratic 2-fold Pfister forms. By assumption,
ϕ1 and ϕ2 are linked. Therefore, the class of ψ in Iq2 (F ) is represented by a form of
dimension 4, hence ψ is isotropic.
(2) ⇒ (1): Let ϕ1 and ϕ2 be two quadratic 2-fold Pfister forms over F . Then ϕ1 − ϕ2 = ψ
for some 6-dimensional form ψ ∈ Iq2 (F ). As ψ is isotropic, we have i0 (ϕ1 ⊥ −ϕ2 ) ≥ 2,
i.e., ϕ1 and ϕ2 are linked.
(1) ⇒ (4): Let Q1 and Q2 be division quaternion algebras over F . Let ϕ1 and ϕ2 be the
reduced norm quadratic forms of Q1 and Q2 respectively. By assumption, ϕ1 and ϕ2 are
linked. In particular, ϕ1 and ϕ2 are split by a separable quadratic field extension L/F .
Hence L splits Q1 and Q2 and therefore L is isomorphic to subfields of Q1 and Q2 .
39. ADDENDUM: LINKED FIELDS AND THE HASSE NUMBER 175
(3) ⇔ (4) ⇔ (5) is proven in Theorem 98.20.

(3) ⇔ (6) is obvious.
(4) ⇒ (1): Let ϕ1 and ϕ2 be two anisotropic 2-fold Pfister forms over F . Let Q1 and Q2
be two division quaternion algebras with the reduced norm forms ϕ1 and ϕ2 respectively.
By assumption, Q1 and Q2 have quadratic subfields isomorphic to a separable quadratic
extension L/F . By Example 9.8, the forms ϕ1 and ϕ2 are divisible by the norm form of
L/F and hence are linked. ¤
A field F is called linked if F satisfies the conditions of Theorem 39.1.

For a formally real field F , the u-invariant can be thought of as a weak Hasse Principle,
i.e., every locally hyperbolic form of dimension > u(F ) is isotropic. A variant of the u-
invariant introduced in [34] naturally arises. We call a quadratic form ϕ over F locally
isotropic or totally indefinite if ϕFP is isotropic at each real closure FP of F (if any) i.e.,
ϕ is indefinite at each real closure of F (if any). The Hasse number of a field F is defined
to be
u
e(F ) := max{dim ϕ | ϕ is a locally isotropic anisotropic form over F }
or ∞ if no such maximum exists. For fields that are not formally real this coincides with
the ū-invariant. If a field is formally real, finiteness of its ue-invariant is a very strong
condition and is a form of a strong Hasse Principle. For example, if F is a global field
then ue(F ) = 4 by Meyer’s Theorem [106] (a forerunner of the Hasse-Minkowski Principle
[125]) and if F is the function field of a real curve then u e(F ) = 2 (cf. Example 39.11
below), but if F/R is formally real and finitely generated of transcendence degree > 1
then, although u(F ) is finite, its Hasse number u e(F ) is infinite.
Exercise 39.2. Show if the Hasse number is finite then it cannot be 3, 5, or 7.
We establish another characterization of ue(F ) using a concept introduced in [42]. We
say F satisfies Property Hn with n > 1 if there exist no anisotropic, locally isotropic forms
of dimension n. Thus if ue(F ) is finite
u
e(F ) + 1 = min{n | F satisfies Hm for all m ≥ n}.
Remark 39.3. Every 6-dimensional form in Iq2 (F ) is locally isotropic, since every
element in I 2 (F ) has signature divisible by 4 at every ordering. Hence if u
e(F ) ≤ 4 then
F is linked by Theorem 39.1.
Lemma 39.4. Let F be a linked field of characteristic not two. Then
(1) Any pair of n-fold Pfister forms are linked for n ≥ 2.
n−3
(2) If ϕ ∈ Pn (F ) then ϕ ¡' hh−w
¢ 1 , xii if n ×= 2 and ϕ = 2 hh−w1 , −w2 , xii if n ≥ 3
for some w1 , w2 ∈ D 3h1i and x ∈ F .
n
(3) For every n ≥ 0 and ϕ ∈ IP (F ), there exists an integer m and ρi ∈ GPi (F ) with
n ≤ i ≤ m satisfying ϕ = m i=n ρi in W (F ). Moreover, if ϕ is a torsion element
then each ρi is torsion.
(4) I 4 (F ) is torsion-free.
Proof. (1), (2): Any pair of n-fold Pfister forms are easily seen to be linked by
induction so (1) is true. As any 2-fold Pfister form is linked to 4h1i, statement (2) holds
for n = 2. Let ρ = hha, b, cii be a 3-fold Pfister form then applying the¡n =¢ 2 case
gives ρ = hhw1 , x, yii = hhw1 , w2 , zii for some x, y, z ∈ F × and w1 , w2 ∈ D 3h1i . This
establishes the n = 3 case. Let ρ = hha, b, c, dii be a 4-fold Pfister form. By assumption,
there exist x, y, z ∈ F × such that hha, bii ' hhx, yii and hhc, dii ' hhx, zii. Thus
(39.5) ρ = hha, b, c, dii ' hhx, y, x, zii ' hh−1, y, x, zii = 2hhy, x, zii.
Statement (2) follows.
(3): Let ψ and τ be n-fold Pfister forms. As they are linked ψ − τ = ahhbii · µ in W (F )
for some (n − 1)-fold Pfister form µ and a, b ∈ F × . Then
xψ + yτ = xψ − xτ + xτ + yτ = axhhbii · µ + xhh−xyii · τ
The first part now follows by repeating this argument. If ϕ is torsion, then inductively,
each ρi is torsion by the Hauptsatz 23.8, so the second statement follows.
(4): By (3), it suffices to show there are no anisotropic torsion n-Pfister forms with
n > 3. By Proposition 35.3, it suffices to show if ρ ∈ P4 (F ) satisfies 2ρ = 0 in W
¡ (F )¢ then
ρ = 0 in W (F ). By Lemma 35.2, we can write ρ ' hha, b, c, wii with w ∈ D 2h1i and
a, b, c ∈ F × . Applying equation (39.5) with d = w, we have ρ ' 2hhy, x, zii ' 2hhy, c, wii
which is hyperbolic. The result follows. ¤
Lemma 39.6. Let char F 6= 2 and n ≥ 2. If F is linked and F satisfies Hn then it
satisfies Hn+1 .
Proof. Let ϕ be an (n + 1)-dimensional anisotropic quadratic form with n ≥ 2.
Replacing ϕ by xϕ for an appropriate x ∈ F × , we may assume that ϕ = hw,¡b, wbi¢ ⊥ ϕ1
for some w, b ∈ F × and form ϕ1 over F and by Lemma 39.4 that w ∈ D 3h1i . Let
ϕ2 = hw, bi ⊥ ϕ1 . As sgnP hbi = sgnP hwbi for all P ∈ X(F ), the form ϕ is locally isotropic
if and only if ϕ2 is. The result follows by induction. ¤
Remark 39.7. If char F 6= 2 and n ≥ 4 then F satisfies Property Hn+1 if it satisfies
Property Hn . However, in general, H3 does not imply H4 (cf. [42]).
Exercise 39.8. Let F be a formally real pythagorean field. Then u e(F ) is finite if and
only if I 2 (F ) = 2I(F ). Moreover, if this is the case then u
e(F ) = 0.
The following theorem from [34] strengthens the result in [37].
Theorem 39.9. Let F be a linked field of characteristic not two. Then u(F ) = u
e(F )
and u
e(F ) = 0, 1, 2, 4, or 8.
Proof. We first show that u e(F ) = 0, 1, 2, 4, or 8. We know that I 4 (F ) is torsion-free
by Lemma 39.4. We first show that F satisfies H9 hence u e(F ) ≤ 8 by Lemma 39.6. Let ϕ
be a 9-dimensional locally isotropic form over F . Replacing ϕ by xϕ for an appropriate
x ∈ F × , we can assume that ϕ = h1i + ϕ1 in W (F ) with ϕ1 ∈ I 2 (F ) using Proposition
4.13. By Lemma 39.4, we have a congruence
(39.10) ϕ ≡ h1i + ρ2 − ρ3 mod I 4 (F )
39. ADDENDUM: LINKED FIELDS AND THE HASSE NUMBER 177
for some ρi ∈ Pi (F ) with i = 2, 3. Write ρ2 ' hha, bii and ρ3 ' hhc, d, eii. As F is linked,
we may assume that e = b and −d ∈ DF (ρ02 ). Thus we have
ϕ ≡ h1i + hha, bii − hhc, d, bii
¡ ¢
≡ h1i − d hha, bii − hhc, bii − hhc, bii
≡ −cdhhac, bii − hhc, bii0 mod I 4 (F ).
Let µ = ϕ ⊥ cdhhac, bii ⊥ hhc, bii0 , a locally isotropic form over F lying in I 4 (F ). In
particular, for all P ∈ X(F ), we have 16 | sgnP µ. As the locally isotropic form µ is
sixteen dimensional, | sgnP µ| < 16 for all P ∈ X(F¡) so sgnP µ ¢ = 0 for all P ∈ X(F ) and
4 0
µ ∈ It (F ) = 0. Consequently, ϕ = −cdhhac, bii ⊥ −hhc, bii in W (F ) so ϕ is isotropic
and ue(F ) ≤ 8.
Suppose that u e(F ) < 8. Then there are no anisotropic torsion 3-fold Pfister forms over
F . It follows that I 3 (F ) is torsion-free by Lemma 39.4. We show u e(F ) ≤ 4. To do this
it suffices to show that F satisfies H5 by Lemma 39.6. Let ϕ be a 5-dimensional, locally
isotropic space over F . Arguing as above but going mod I 3 (F ), we may assume that
ϕ ≡ h1i − hha, bii = −hha, bii0 mod I 3 (F ).
Let µ = ϕ ⊥ hha, bii0 , an 8-dimensional, locally isotropic form over F lying in I 3 (F ). As
above, it follows that µ is locally hyperbolic hence µ ∈ It3 (F ) = 0. Thus ϕ = −hha, bii0 in
W (F ) so isotropic and ue(F ) < 4. In a similar way, we see that u e(F ) = 0, 1, 2 are the only
other possibilities. This shows that ue(F ) = 0, 1, 2, 4, 8. The argument above and Lemma
39.4 show that u(F ) = u e(F ). ¤
Note the proof shows if F is linked and I n (F ) is torsion-free then u
e(F ) ≤ 2n−1 .
It is still unknown whether Theorem 39.9 is true if char F = 2.
√ ¡ √ ¢
Example 39.11. (1) If F ( −1) is a C1 field then I 2 F ( −1) = 0. It follows that
I 2 (F ) = 2I(F ) and is torsion free by Corollary 35.14 and Proposition 35.1 (or Corollary
35.27). In particular, F is linked and u e(F ) ≤ 2.
(2) If F is a local or global field then u
e(F ) = 4.
(3) Let F0 be a local field and F = F0 ((t)) be a Laurent series field. As u(F0 ) = 4, and
F is not formally real, we have u e(F ) = u(F ) = 8. If char F 6= 2, this field F is linked by
the following exercise:
Exercise 39.12. Let F = F0 ((t)) with char F 6= 2. Show there exist no 4-dimensional
anisotropic spaces of discriminant different from F0 ×2 over F0 if and only if F is linked.
There exist linked formally real fields with Hasse number 8, but the construction of
such fields is more delicate (cf. [44]).
Remark 39.13. Let F be a formally real field. Then it can be shown that u e(F ) is
2 n
finite if and only if u(F ) if finite and I (Fpy ) = 2I (Fpy ) (cf. [44]). If both of these
invariants are finite, they may be different (cf. [116].)
Remark¡ √39.14.
¢ Wensay that a formally real field
√ F satisfies condition Sn if the map
n
s∗¡ : Ired ¢ F ( w) → Ired (F ) induced by s : F ( w) → F with s(1) = 0 for all w ∈
D ∞h1i is surjective (cf. Exercise 35.4). In [44], it is shown that u
e(F ) is finite if and
only if u(F ) is finite, F satisfies S1 , and str (F ) ≤ 1. Moreover, it is also shown in [44]
that str (Fpy ) ≤ n if and only if str (F ) ≤ n and F satisfies Sn . This last statement is the
real analog of the first three conditions of Corollary 35.27.
CHAPTER VII
Applications of the Milnor Conjecture
40. Exact sequences for quadratic extensions
In this section, we derive the first consequences of the validity of the Milnor Conjecture
for fields of characteristic different from two. In particular, we show that the infinite
complexes of the powers of I- (cf. 34.20) and I- ¯ (cf. 34.21) arising from a quadratic
extension of a field of characteristic different from two are in fact exact as shown in the
paper [110] of Orlov-Vishik-Voevodsky. For fields of characteristic two, we also show
this to be true for separable quadratic extensions as well as proving the exactness of the
corresponding complexes (34.27) and (34.28) for purely inseparable quadratic extensions.
In addition, we show that for all fields, the ideals Iqn (F ) coincide with the ideals Jn (F )
defined in terms of the splitting patterns of quadratic forms.
We need the following lemmas.
Lemma 40.1. Let K/F be a quadratic field extension and s : K → F a nonzero
F -linear functional such that s(1) = 0. Then for every n ≥ 0, the diagram
cK/F
kn (K) −−−→ kn (F )
 

fn y
f
yn
n ∗ s n
I (K) −−−→ I (F )
commutes where the vertical homomorphisms are defined in (5.1).
Proof. As all the maps in the diagram are K∗ (F )-linear, in view of Lemma 34.16, it
is sufficient to check commutativity only when n = 1. The statement follows now from
Corollary 34.19 and Fact 100.8. ¤
Lemma 40.2. Suppose that F is a field of characteristic 2 and K/F a quadratic field
extension. Let s : K → F be a nonzero F -linear functional satisfying s(1) = 0. Then the
diagram
n s∗ n
I q (K) −−− → I q (F )
 

ēn y
ē
yn
cK/F
H n (K) −−−→ H n (F )
is commutative.
Proof. It follows from Lemma 34.16 that it is sufficient to prove the statement in the
case n = 1. This follows from Lemma 34.14, since the corestriction map cK/F : H 1 (K) →
H 1 (F ) is induced by the trace map TrK/F : K → F (cf. Example 101.2). ¤
179
180 VII. APPLICATIONS OF THE MILNOR CONJECTURE
We set I n (F ) = W (F ) if n ≤ 0.
We first consider the case char F 6= 2.
√Theorem 40.3. Suppose that F is a field ×

of characteristic different from 2 and K =
F ( a)/F a quadratic extension with a ∈ F . Let s : K → F be an F -linear functional
such that s(1) = 0. Then the following infinite sequences
s ·hhaii rK/F s ·hhaii
··· −
→∗
I n−1 (F ) −−−→ I n (F ) −−−→ I n (K) −
→∗
I n (F ) −−−→ I n+1 (F ) → · · · ,
∗ s n−1 ·hhaii n rK/F

∗ n s n ·hhaii n+1
··· −
→ I (F ) −−−→ I (F ) −−−→ I (K) −
→ I (F ) −−−→ I (F ) → · · ·
are exact.
Proof. Consider the diagram
·{a} rK/F cK/F ·{a}
kn−1 (F ) −−−→ kn (F ) −−−→ kn (K) −−−→ kn (F ) −−−→ kn+1 (F )
    
    
y y y y y
n−1 ·hhaii n rK/F
∗ n s n ·hhaii n+1
I (F ) −−−→ I (F ) −−−→ I (K) −−− → I (F ) −−−→ I (F )
where the vertical homomorphisms are defined in (5.1). It follows from Lemma 40.1
that the diagram is commutative. By Fact 5.15, the vertical maps in the diagram are
isomorphisms. The top sequence in the diagram is exact by Proposition 101.10. Therefore,
the bottom sequence is also exact.
To prove exactness of the first sequence in the statement consider the commutative
diagram
I n+1 (F ) → I n+1 (K)→ I n+1 (F )→ I n+2 (F )→ I n+2 (K)
    
    
y y y y y
I n−1 (F )→ I n (F ) → I n (K) → I n (F ) → I n+1 (F )→ I n+1 (K)
    
    
y y y y y
n−1 n n n n+1
I (F )→ I (F ) → I (K) → I (F ) → I (F )
with the horizontal sequences considered above and natural vertical maps. By the first
part of the proof the bottom sequence is exact. Therefore, exactness of the middle se-
quence implies exactness of the top one. Thus the statement follows by induction on n
(with the base of the induction given by Corollary 34.12). ¤
Remark 40.4. Let char F 6= 2. Then the second exact sequence in Theorem 40.3 can
be rewritten as
GW (K)
r9 LLL
rrrrr
rK/F LLLs∗
LLL
r
rrr L%
GW (F ) o GW (F )
·hhaii
is exact (cf. Corollary 34.12).

40. EXACT SEQUENCES FOR QUADRATIC EXTENSIONS 181
Now consider the case of fields of characteristic 2. We consider separately the cases of
separable and purely inseparable quadratic field extensions.
Theorem 40.5. Suppose that F is a field of characteristic 2 and K/F a separable
quadratic field extension. Let s : K → F be a nonzero F -linear functional such that
s(1) = 0. Then the following sequences
rK/F s · NK/F rK/F s
0 → I n (F ) −−−→ I n (K) −
→∗
I n (F ) −−−−→ Iqn+1 (F ) −−−→ Iqn+1 (K) −
→∗
Iqn+1 (F ) → 0,
n rK/F ∗ n s n · NK/F n+1 rK/F n+1 ∗ s n+1

0 → I (F ) −−−→ I (K) −
→ I (F ) −−−−→ I q (F ) −−−→ I q (K) −
→ Iq (F ) → 0
are exact.
rK/F cK/F ·[K] rK/F cK/F
0 → kn (F ) −−−→ kn (K) −−−→ kn (F ) −−−→ H n+1 (F ) −−−→ H n+1 (K) −−−→ H n+1 (F ) → 0
   x x x
     
y y y   
n rK/F ∗ n s n · NK/F n+1 rK/F n+1 ∗ s n+1
0 → I (F ) −−−→ I (K) −−−→ I (F ) −−−−→ I q (F ) −−−→ I q (K) −−−→ Iq (F ) → 0
where the vertical homomorphisms are defined in (5.1) and Fact 16.2. The middle map
in the top row is the multiplication by the class [K] ∈ H 1 (F ). By Proposition 101.12,
the top sequence is exact. By Facts 5.15 and 16.2, the vertical maps are isomorphisms.
Therefore, the bottom sequence is exact.
Exactness of the second sequence in the statement follows by induction on n from the
first part of the proof and commutativity of the diagram
0 → I n+1 (F )→ I n+1 (K)→ I n+1 (F )→ Iqn+2 (F )→ Iqn+2 (K) → Iqn+2 (F ) → 0
     
     
y y y y y y
0 → I n (F ) → I n (K) → I n (F ) → Iqn+1 (F )→ Iqn+1 (K) → Iqn+1 (F ) → 0
     
     
y y y y y y
n n n n+1 n+1 n+1
0 → I (F ) → I (K) → I (F ) → I q (F )→ I q (K)→ I q (F ) → 0.
The base of the induction follows from Corollary 34.15. ¤
Theorem 40.6. Suppose that F is a field of characteristic 2 and K/F a purely in-
separable quadratic field extension. Let s : K → F be an F -linear functional such that
s(1) = 0. Then the following sequences
s rK/F s rK/F s
··· −
→∗
I n (F ) −−−→ I n (K) −
→∗
I n (F ) −−−→ I n (K) −
→∗
··· ,
∗ s n ∗
rK/F n ∗s n rK/F n s
··· −
→ I (F ) −−−→ I (K) −
→ I (F ) −−−→ I (K) −
→ ··· ,
are exact.

rK/F cK/F rK/F
kn (F ) −−−→ kn (K) −−−→ kn (F ) −−−→ kn (K)
   
   
y y y y
n rK/F n ∗ s n rK/F n
I (F ) −−−→ I (K) −−−→ I (F ) −−−→ I (K)
where the vertical homomorphisms are defined in (5.1). The diagram is commutative by
Lemma 40.1. By Fact 5.15, the vertical maps in the diagram are isomorphisms. The top
sequence in the diagram is exact by Proposition 100.12. Therefore, the bottom sequence is
also exact. The proof of exactness of the second sequence in the statement of the theorem
is similar to the one in Theorems 40.3 and 40.5. ¤
For further applications, we shall also need the following result which follows immedi-
ately from the validity of the Milnor Conjecture proven by Voevodsky in [142] and Orlov,
Vishik, and Voevodsky in [110, Th. 2.1]:
Fact 40.7. Let char F 6= 2 and ρ a quadratic n-fold Pfister form over F . Then the
sequence
a P
cL/F ∪en (ρ) rF (ρ)/F ¡ ¢
H (L) −−−−→ H ∗ (F ) −−−−→ H ∗+n (F ) −−−−→ H ∗+n F (ρ) ,
∗
is exact, where the direct sum is taken over all quadratic field extensions L/F such that
ρL is isotropic.
Fact 40.8. ([13, Th. 5.4]) Let char F = 2 and ¡let ρ be
¢ a quadratic n-fold Pfister form
n n
over F . Then the kernel of rF (ρ)/F : H (F ) → H F (ρ) coincides with {0, en (ρ)}.
Corollary 40.9. Let ρ be a quadratic n-fold Pfister form over
¢ an arbitrary field F .
n n¡
Then the kernel of the natural homomorphism I q (F ) → I q F (ρ) coincides with {0, ρ̄}.
n ∼
Proof. Under the isomorphism I q (F ) → H n (F ) (cf. Fact 16.2), the homomorphism
¡ ¢
in the statement is identified with H n (F ) → H n F (ρ) . The statement now follows from
Fact 40.7 if char F 6= 2 and Fact 40.8 if char F = 2. ¤
The following statement generalizes Proposition 25.13 and was first proven by Orlov,
Vishik, and Voevodsky in [110].
Theorem 40.10. If F is a field then Jn (F ) = Iqn (F ) for every n ≥ 1.
Proof. By Corollary 25.12, we have the inclusion Iqn (F ) ⊂ Jn (F ). Let ϕ ∈ Jn (F ).

We show by induction on n that ϕ ∈ Iqn (F ). As ϕ ∈ Jn−1 (F ), by the induction hypothesis,
we have ϕ ∈ Iqn−1 (F ). The form ϕ is a sum of m general (n − 1)-fold Pfister forms in
Iqn−1 (F ) for some m. Let ρ be one of them. Let K = F (ρ). Since ϕK is a sum of m − 1
general (n − 1)-Pfister forms in Iqn−1 (K), by induction on m we have ϕK ∈ Iqn (K). By
Corollary 40.9, we have either ϕ ∈ Iqn (F ) or ϕ ≡ ρ modulo Iqn (F ). But the latter case
does not occur as ϕ ∈ Jn (F ) and ρ ∈/ Jn (F ). ¤
41. ANNIHILATORS OF PFISTER FORMS 183
41. Annihilators of Pfister forms
The main purpose of this section is to establish the generalization of Corollary 6.23
and Theorem 9.13 on the annihilators of bilinear and quadratic Pfister forms and show
these annihilators respect the grading induced by the fundamental ideal. We even show
if α is a bilinear or quadratic Pfister form then the annihilator annW (F ) (α) ∩ I n (F ) is
not only generated by bilinear Pfister forms annihilated by α but is, in fact, generated
by bilinear n-fold Pfister forms of the type b ⊗ c with b ∈ annW (F ) (α) a 1-fold bilinear
Pfister form and c a bilinear (n − ¡1)-fold¢Pfister form. In particular, Pfister forms of the
type hhw, a2 , . . . , an ii with w ∈ D ∞h1i and ai ∈ F × generate Itn (F ), thus solving the
problems raised at the end of §33. These results first appeared in [9].
Let F be a field. The smallest integer n such that I n+1 (F ) = 2I n (F ) and I n+1 (F ) is
torsion free is called the stable range of F and is denoted by st(F ). We say that F has
finite stable range if such an n exists and write st(F ) = ∞ if such an n does not exist.
By Corollary 35.30, a field F has finite stable range if and only if I n+1 (F ) = 2I n (F ) for
some integer n. If F is not formally real then st(F ) is the smallest integer n such that
I n+1
¡ (F√) = ¢0. If F is formally real then it follows from Corollary¡ √35.27¢ that st(F ) =
n+1
st F ( −1) , i.e., st(F ) is the smallest integer n such that I F ( −1) = 0.
Lemma 41.1. Suppose that F has finite transcendence degree n over its prime subfield.
Then st(F ) ≤ n + 2 if char F = 0 and st(F ) ≤ n + 1 if char F > 0.
Proof. If the characteristic of F is positive then F is a Cn+1 -field (cf. 97.7) as
finite fields are C1 fields by the Chevalley-Warning Theorem (cf. [127, I.2, Th. 3]) and,
therefore, every (n+2)-fold Pfister form is isotropic, so I n+2 (F ) = 0, i.e., √
st(F ) ≤ n+1. If
the characteristic of F is zero then the cohomological 2-dimension of F ( −1) is at most
n + 2 by [128, II.4.1,¡ √ Prop.¢ 10 and II.4.2, Prop. 11]. By Fact 16.2 and the Hauptsatz
n+3
23.8, we have I F ( −1) = 0. Thus st(F ) ≤ n + 2. ¤
As many problems in a field F reduce to finitely many elements over its prime field,
we can often reduce to a problem over a given field to another over a field having finite
stable range. We then can try to solve the problem when the stable range is finite. We
shall use this idea repeatedly below.
Exercise 41.2. Let K/F be a finite simple extension of degree r. If I n (F ) = 0 then
I n+r (K) = 0. In particular, if a field has finite stable range then any finite extension also
has finite stable range.
Next we study graded annihilators.

Let b be a bilinear n-fold Pfister form. For any m ≥ 0 set
annm (b) = {a ∈ I m (F ) | a · b = 0 ∈ W (F )},
m
annm (b) = {ā ∈ I (F ) | ā · b̄ = 0 ∈ GW (F )}.
Similarly, for a quadratic n-fold Pfister form ρ and any m ≥ 0 set
annm (ρ) = {a ∈ I m (F ) | a · ρ = 0 ∈ Iq (F )},
m
annm (ρ) = {ā ∈ I (F ) | ā · ρ̄ = 0 ∈ I q (F )}.
It follows from Corollary 6.23 and Theorem 9.13 that ann1 (b) and ann1 (ρ) are gener-
ated by the binary forms in them. Thus the following theorem determines completely the
graded annihilators.
Theorem 41.3. Let b and ρ be bilinear and quadratic n-fold Pfister forms respectively.
Then for any m ≥ 1, we have
m−1
annm (b) = I m−1 (F ) · ann1 (b), annm (b) = I (F ) · ann1 (b),
m−1
annm (ρ) = I m−1 (F ) · ann1 (ρ), annm (ρ) = I (F ) · ann1 (ρ).
Proof. The case char F = 2 is proven in [14, Th.1.1 and 1.2]. We assume that
char F 6= 2. It is sufficient to consider the case of the bilinear form b.
It follows from Fact 40.7 and Fact 16.2 that the sequence
a m P
s∗ m ·b n+m
I (L) −−→ I (F ) − →I (F ),
is exact where the direct sum is taken over all quadratic field extensions L/F with bL
isotropic. By Lemma 34.16, we have I m (L) = I m−1 (F )I(L) hence the image of s∗ :
I m (L) → I m (F ) is contained in I m−1 (F ) · ann1 (b). Therefore, the kernel of the second
m
map in the sequence coincides with the image of I m−1 (F ) · ann1 (b) in I (F ). This proves
m−1
annm (b) = I (F ) · ann1 (b).
Let c ∈ annm (b). We need to show that c ∈ I m−1 (F ) · ann1 (b). We may assume that
F is finitely generated over its prime field and hence F has finite stable range by Lemma
41.1. Let k be an integer such that k + m > st(F ). Repeatedly applying exactness
of the sequence above, we see that c is congruent to an element a ∈ I k+m (F ) modulo
I m−1 (F ) · ann1 (b). Replacing c by a we may assume that m > st(F ).
We claim that it suffices to prove the result for c an m-fold Pfister form. By Theorem
33.15, for any c ∈ I m (F ), there is an integer n such that
r
X
2n sgn c = ki · sgn ci ,
i=1
with ki ∈ Z and (n + m)-fold Pfister forms ci with pairwise disjoint supports. As m >
st(F ), it follows from Proposition 35.22 that ci ' 2n di for some m-fold Pfister forms di .
Since I m (F ) is torsion free, we have
r
X
c= ki · di
i=1
m
in I (F ) and the supports of the di ’s are pairwise disjoint. In particular, if b ⊗ c is
hyperbolic then supp(b) ∩ supp(c) = ∅, so supp(b) ∩ supp(di ) = ∅ for every i. As I m (F )
is torsion free, this would mean that b ⊗ ci is hyperbolic for every i and establishes the
claim. Therefore, we may assume that c is a Pfister form.
The result now follows from Lemma 35.18(1). ¤
41. ANNIHILATORS OF PFISTER FORMS 185
We turn to the generators for Itn (F ), the torsion in I n (F ).

Theorem 41.4. For any field F we have Itn (F ) = I n−1 (F )It (F ).
Proof. Let c ∈ Itn (F ). Then 2m c = 0 for some m. Applying Theorem 41.3 to the
Pfister form b = 2m h1i, we have
c ∈ annn (b) = I n−1 (F ) · ann1 (b) ⊂ I n−1 (F )It (F ). ¤
Recall that by Proposition 31.30, the group It (F ) is generated by binary torsion forms.
Hence Theorem 41.4 yields
Corollary 41.5. A field F satisfies property An if and only if I n (F ) is torsion-free.
Remark 41.6. By Theorem 41.4, every torsion bilinear n-fold Pfister form b can
¢ as a Z-linear combination of the (torsion) forms hha1 , a2 , . . . , an ii with a1 ∈
be¡ written
D ∞h1i . Note that b itself may not be isometric to a form like this (cf. Example 32.4).
Theorem 41.7. Let b and ρ be bilinear and quadratic n-fold Pfister forms respectively.
Then for any m ≥ 0, we have
W (F )b ∩ I n+m (F ) = I m (F )b,
W (F )ρ ∩ Iqn+m (F ) = I m (F )ρ.
Proof. We prove the first equality (the second being similar). Let c ∈ W (F )b ∩
n+m
I (F ). We show by induction on m that c ∈ I m (F )b. Suppose that c = a · b in W (F )
for some a ∈ I m−1 (F ), i.e., ā ∈ annm−1 (b). By Theorem 41.3, we have ā = d̄ē for some
d ∈ I m−2 (F ) and e ∈ W (F ) satisfying ē ∈ ann1 (b). Let f be a binary bilinear form
n+1
congruent to e modulo I 2 (F ). As f̄b̄ = ēb̄ = 0 ∈ I (F ), the general (n + 1)-fold Pfister
form f ⊗ b belongs to I n+2 (F ). By the Hauptsatz 23.8, we have f · b = 0 in W (F ). Since
a ≡ df modulo I m (F ) it follows that c = ab ∈ I m (F )b. ¤
Exercise 41.8. Let b and c be bilinear k-fold and n-fold Pfister forms respectively
over a field F of characteristic not 2. Prove that for any m ≥ 1 the group
W (F )c ∩ annW (F ) (b) ∩ I m+n (F )
is generated by (m + n)-fold Pfister forms d in annW (F ) (b) that are divisible by c.
The theorem allows us to answer the problems at the end of §33. The solution to
Lam’s problem in [89] was proven in [9] and independently by Dickmann and Miraglia
(cf. [30]).
Corollary 41.9. Let b be a form over F . If 2n b ∈ I n+m (F ) then b ∈ I m (F )+Wt (F ).
In particular,
¡ ¢
sgn(b) ∈ C X(F ), 2m Z if and only if b ∈ I m (F ) + Wt (F ).
¡ ¢
Proof. Suppose that sgn b ∈ C X(F ), 2m Z . By Theorem 33.15, there exists a form
a ∈ I n+m (F ) such that 2n sgn b = sgn a for some n. In particular, 2n b − a ∈ Wt (F ).
Therefore 2k+n b = 2k a for some k. By Theorem 41.7, applied to the form 2k+n h1i, we
may write 2k a = 2k+n c for some c ∈ I n (F ). Then b − c lies in Wt (F ) as needed. ¤
As a consequence, we obtain a solution to Marshall’s question in [98]. It was first

proved by Dickmann and Miraglia in [31].
Corollary 41.10. Let F be a formally real pythagorean ¡ field.
¢ Let
¡ b be a form¢ over
F . If 2n b ∈ I n+m (F ) then b ∈ I m (F ). In particular, sgn I m (F ) = C X(F ), 2m Z .
¡ ¢
If F is a formally real let GC X(F ), Z be the graded ring
¡ ¢ a ¡ ¢ ¡ ¢ a ¡ ¢
GC X(F ), Z := 2n C X(F ), Z /2n+1 C X(F ), Z = C X(F ), 2n Z/2n+1 Z
and GWt (F ) the graded ideal in GW (F ) induced by It (F ). Then Corollary 41.9 implies
that the signature induces an exact sequence
¡ ¢
0 → GWt (F ) → GW (F ) → GC X(F ), Z
and Corollary 41.10 says if F is a formally
¡ real
¢ pythagorean field then the signature
induces an isomorphism GW (F ) → GC X(F ), Z .
We interpret this result in terms of the reduced Witt ring and prove the result men-
tioned at the end of §34.
Theorem 41.11. Let K be a quadratic extension and s : K → F a nonzero F -linear
functional such that s(1) = 0. Then the sequence
rK/F s
n n n ∗ n
0 → Ired (K/F ) → Ired (F ) −−−→ Ired (K) −
→ Ired (F )
is exact.
n n
Proof. We need only to show exactness at Ired (K). Let c ∈ Ired (K) satisfy sP∗ (c) is
n
trivial in Ired (F ), i.e., the form s∗ (c) is torsion. By Theorem 41.4, we have s∗ (c) = ai bi
n−1
with ai ∈ I (F ) and bi ∈ It (F ). It follows by Corollary P 34.32 that bi = s∗ (di ) for some
torsion forms di ∈ I(K). Therefore, the form e := c − (ai )K di belongs to the kernel of
s∗ : I n (K) → I n (F ). It follows from Theorems 40.3 and 40.5 that e = rK/F (f) for some
f ∈ I n (F ). Therefore c ≡ rK/F (f) modulo torsion. ¤
42. Presentation of I n (F )
In this section, using the validity of the Milnor Conjecture, we show that the presen-
tation established for I 2 (F ) in Theorem 4.22 generalizes to a presentation for I n (F ). This
was first proven in [9] in the case of characteristic not two. The characteristic two case
was also proven independently by Arason and Baeza from that given below in [3].
Let n ≥ 2 and let I n (F ) be the abelian group generated by all the isometry classes [b]
of bilinear n-fold Pfister forms b subject to the generating relations:
£ ¤
(1) hh1, 1, . . . , 1ii = 0.
£ ¤ £ ¤ £ ¤ £ ¤
(2) hhab, cii ⊗ d + hha, bii ⊗ d = hha, bcii ⊗ d + hhb, cii ⊗ d for all a, b, c ∈ F ×
and bilinear (n − 2)-fold Pfister forms d.
42. PRESENTATION OF I n (F ) 187
Note that the group I 2 (F ) was defined earlier in Section §4.C.

There is a natural surjective group homomorphism gn : I n (F ) → I n (F ) taking the
class [b] of a bilinear n-fold Pfister form b to b in I n (F ). The map g2 is an isomorphism
by Theorem 4.22.
£ As in the proof ¤ of Lemma 4.18, applying both relations repeatedly, we find that
hha1 , a2 , . . . , an ii = 0 if a1 = 1. It follows that for any bilinear m-fold Pfister form b,
the assignment a 7→ a ⊗ b gives rise to a well defined homomorphism
I n (F ) → I n+m (F )
taking [a] to [a ⊗ b].
Lemma 42.1. Let b be a metabolic bilinear n-fold Pfister form. Then [b] = 0 in I n (F ).
Proof. We prove the statement by induction on n. Since g2 is an isomorphism, the
statement is true if n = 2. In the general case, we write b = hhaii ⊗ c for some a ∈ F × and
bilinear (n − 1)-fold Pfister form c. We may assume by induction that c is anisotropic. It
follows from Corollary 6.14 that c ' hhbii ⊗ d for some b ∈ F × and bilinear£ (n − 2)-fold
¤
Pfister form d such that hha, bii is metabolic. By the case n = 2, we have hha, bii = 0
in I 2 (F ), hence [b] = [hha, bii ⊗ d] = 0 in I n (F ). ¤
For each n, let αn : I n+1 (F ) → I n (F ) be the map given by
[hha, bii ⊗ c] 7→ [hhaii ⊗ c] + [hhbii ⊗ c] − [hhabii ⊗ c].
We show that this map is well defined. Let hha, bii ⊗ c and hha0 , b0 ii ⊗ c0 be isometric
bilinear n-fold Pfister forms. We need to show that
(42.2) [hhaii ⊗ c] + [hhbii ⊗ c] − [hhabii ⊗ c] = [hha0 ii ⊗ c0 ] + [hhb0 ii ⊗ c0 ] − [hha0 b0 ii ⊗ c0 ]
in I n (F ). By Theorem 6.10, the forms hha, bii ⊗ c and hha0 , b0 ii ⊗ c0 are chain p-equivalent.
Thus we may assume that one of the following cases hold:
(1) a = a0 , b = b0 and c ' c0 .
(2) hha, bii ' hha0 , b0 ii and c = c0 .
(3) a = a0 , c = hhcii⊗d, and c0 = hhc0 ii⊗d for some c ∈ F × and bilinear (n−2)-fold
Pfister form d and hhb, cii ' hhb0 , c0 ii.
It follows that it is sufficient to prove the statement in the case n = 2. The equality
(42.2) holds if we compose the morphism α2 with the homomorphism g2 : I 2 (F ) → I 2 (F ).
But g2 is an isomorphism, hence αn is well defined. It is easy to check that αn is a
homomorphism.
The homomorphism αn fits in the commutative diagram
nα
I n+1 (F ) −−−→ I n (F )
 

gn+1 y
 gn
y
I n+1 (F ) −−−→ I n (F )
with the bottom map the inclusion.
Lemma 42.3. The natural homomorphism

n
γ : coker(αn ) → I (F )
is an isomorphism.
Proof. Consider the map
£ ¤
τ : (F × )n → coker(αn ) given by (a1 , a2 , . . . , an ) 7→ hha1 , a2 , . . . , an ii + im(αn ).
Clearly τ is symmetric with respect to permutations of the ai ’s.
By definition of αn we have
£ ¤ £ ¤ £ ¤
hhaii ⊗ c + hhbii ⊗ c ≡ hhabii ⊗ c mod im(αn )
for any bilinear (n − 1)-fold Pfister form c. It follows that τ is multilinear.
£ ¤
The map τ also satisfies the Steinberg relation. £Indeed if a1 +a2 =¤ 1, then hha1 , a2 ii =
0 in I 2 (F ) as g2 is an isomorphism and therefore hha1 , a2 , . . . , an ii = 0 in I n (F ).
As the group coker(αn ) has exponent 2, the map τ induces a group homomorphism
kn (F ) = Kn (F )/2Kn (F ) → coker(αn )
which we also denote by τ . The composition γ ◦ τ takes a symbol {a1 , a2 , . . . , an } to
hha1 , a2 , . . . , an ii + I n+1 (F ). By Fact 5.15, the map γ ◦ τ is an isomorphism. As τ is
surjective, we have γ is an isomorphism. ¤
It follows from Lemma 42.3 that we have a commutative diagram
α
n n
I n+1 (F ) −−−→ I n (F ) −−−→ I (F ) −−−→ 0
  °

gn+1 y

gn y
°
°
n
0 −−−→ I n+1 (F ) −−−→ I n (F ) −−−→ I (F ) −−−→ 0
with exact rows. It follows that if gn+1 is an isomorphism then gn is also an isomorphism.
Theorem 42.4. If n ≥ 2, the abelian group I n (F ) is generated by the isometry classes
of bilinear n-fold Pfister forms subject to the generating relations
(1) hh1, 1, . . . , 1ii = 0.
(2) hhab, cii · d + hha, bii · d = hha, bcii · d + hhb, cii · d for all a, b, c ∈ F × and bilinear
(n − 2)-fold Pfister forms d.
Proof. We shall show that the surjective map gn : I n (F ) → I n (F ) is an isomor-
phism. Any element in the kernel of gn = gn,F belongs to the image of the natural map
gn,F 0 → gn,F where F 0 is a subfield of F , finitely generated over the prime subfield of F .
Thus we may assume that F is finitely generated. It follows from Lemma 41.1 that F has
finite stable range. The discussion preceding the theorem shows that we may also assume
that n > st(F ).
If F is not formally real then I n (F ) = 0, i.e., every bilinear n-fold Pfister form is
metabolic. By Lemma 42.1, the group I n (F ) is trivial and we are done.
In what follows we may assume that F is formally real, in particular, char F 6= 2.
We let M be the abelian group given by generators {b}, the isometry classes of bilinear
n-fold Pfister forms b over F , and relations {b} = {c} + {d} where the bilinear n-fold
42. PRESENTATION OF I n (F ) 189
Pfister forms b, c and d satisfy b = c + d in W (F ). In particular, {b} = 0 in M if b = 0

in W (F ).
We claim that the homomorphism
δ : M → I n (F ) given by {b} 7→ [b]
is well defined. To see this, it suffices to check that if b, c and d satisfy b = c + d in W (F )
then [b] = [c] + [d] in I n (F ). As char F 6= 2, it follows from Proposition 24.5 that there
are c, d ∈ F × and a bilinear (n − 1)-fold Pfister form a such that
c ' hhcii ⊗ a, d ' hhdii ⊗ a, b ' hhcdii ⊗ a.
The equality b = c + d implies that hhc, dii · a = 0 in W (F ). Therefore,
¡ ¢
0 = αn [hhc, dii ⊗ a] = [c] + [d] − [b]
in I n (F ), hence the claim.
Let b be a bilinear n-fold Pfister form and d ∈ F × . As I n+1 (F ) = 2I n (F ), we can
write hhdii · b = 2c and hh−dii · b = 2d in W (F ) with c, d bilinear n-fold Pfister forms.
Adding, we then get 2b = 2c + 2d in W (F ), hence b = c + d since I n (F ) is torsion free.
It follows that [b] = [c] + [d] in M . We generalize this as follows:
Lemma 42.5. Let F be a formally real field having finite stable range. Suppose that
n is a positive integer ≥ st(F ). Let b ∈ Pn (F ) and d1 , . . . , dm ∈ F × . For every ε =
m m
(ε
P1 , . . . , εm ) ∈ {±1} write hhε1 d1 , . . . , εm dm ii ⊗ b ' 2 cε with cε ∈ Pn (F ). Then [b] =
ε [cε ] in M .
Proof. We induct on m: The case m = 1 was done above. So we assume that m > 1.
For every ε0 = (ε2 , . . . , εm ) ∈ {±1}m−1 write
hhε2 d2 , . . . , εm dm ii ⊗ b ' 2m−1 dε0
P
with dε0 ∈ Pn (F ). By the induction hypothesis, we then have [b] = ε0 [dε0 ] in M . It
therefore suffices to show that [dε0 ] = [c(1,ε0 ) ] + [c(−1,ε0 ) ] for every ε0 . But
2m dε0 = 2hhε2 d2 , . . . , εm dm ii · b
¡ ¢
= hhd1 ii + hh−d1 ii · hhε2 d2 , . . . , εm dm ii · b
= 2m c(1,ε0 ) + 2m c(−1,ε0 )
in W (F ) hence dε0 = c(1,ε0 ) + c(−1,ε0 ) in W (F ). Consequently, [dε0 ] = [c(1,ε0 ) ] + [c(−1,ε0 ) ] in
M. ¤
Proposition 42.6. Let F be a formally real field having finite stable range. Suppose
Ps integer ≥ st(F ). Then every element in M can be written as a Z-linear
that n is a positive
combination j=1 lj · [cj ] with forms c1 , . . . , cs ∈ Pn (F ) having pairwise disjoint supports
in X(F ).
Pr
Proof. Let a = i=1 ki · [bi ] ∈ M . Write bi ' hhai1 , . . . , ain ii for i = 1, . . . , r.
Nr Nn
For every matrix ε = (εik )r,n
i=1,k=1 in {±1}
r×n
let fε ' j=1 l=1 hhεjl ajl ii and write
fε ⊗ bi ' 2rnP
ci,ε with ci,ε bilinear n-fold Pfister forms for i = 1, . . . , r. By Lemma 42.5, we
have [bi ] = ε [ci,ε ] in M for i = 1, . . . , r, hence
r
X r
X X r
XX
a= ki · [bi ] = ki · [ci,ε ] = ki · [ci,ε ]
i=1 i=1 ε ε i=1
in M .
For each ε, write fε ' 2nr−n dε with dε a bilinear n-fold Pfister form. Clearly, the fε
have pairwise disjoint supports, hence also the dε . Now look at a pair (i, ε). If all the
εik , k = 1, . . . , r, are 1 then fε ⊗ bi = 2n fε = 2nr dε hence ci,ε = dε . If, however, some εik ,
k
P= 1, . . . , r, is −1 then fε ⊗ bi = 0 hence ci,ε = 0. It follows that for each ε we have
r
i=1 ki · [ci,ε ] = lε · dε for some integer lε . Consequently,
r
XX X
a= ki · [ci,ε ] = l ε · dε . ¤
ε i=1 ε
Applying Proposition 42.6 to an element in the kernel of the composition

gn sgn ¡ ¢
→ I n (F ) −→ I n (F ) −−→ C X(F ), Z ,
δ
M−
we see that all the coefficients lj are 0. Hence the composition is injective. Since δ is
surjective, it follows that gn is injective and therefore an isomorphism. The proof of
Theorem 42.4 is complete. ¤
Remark 42.7. In [3], Arason and Baeza also establish a presentation of Iqn (F ). They
show that as an additive group Iqn (F ) for n ≥ 2 is isomorphic to the equivalence classes
{ϕ} of n-fold quadratic Pfister forms ϕ over F subject to the relations:
© ª © ª © ª
(1q ) α ⊗ [1, d1 + d2 ] = α ⊗ [1, d1 ] + α ⊗ [1, d2 ] for any (n − 1)-fold bilinear
Pfister form
© ª α© and d1 , ª
d2 ∈©F . ª © ª
(2q ) hhaii⊗ρ + hhbii⊗ρ = hha+bii⊗ρ + hhab(a+b)ii⊗ρ for any (n−1)-fold
quadratic Pfister form ρ and a, b ∈ F × satisfying a + b 6= 0.
43. Going down and torsion-freeness
We show in this section that if K/F is a finite extension with I n (K) torsion free then
I n (F ) is torsion free. Since we already know this to be true if char F = 2 by Lemma
35.5, we need only show this when char F 6= 2. In this case, we show that the solution of
the Milnor Conjecture that the norm residue map is an isomorphism implies the result as
shown in [8] and, in fact, leads to the more general Corollary 43.10 below.
Let F be a field of characteristic not 2. For any integer k, n ≥ 0 consider Galois
cohomology groups (cf. §101)
H n (F, k) := H n,n−1 (F, Z/2k Z).
In particular H n (F, 1) = H n (F ).
By Corollary 101.7, there is an exact sequence
0 → H n (F, r) → H n (F, r + s) → H n (F, s).
43. GOING DOWN AND TORSION-FREENESS 191
For a field extension L/F set

¡ rL/F ¢
H n (L/F, k) := ker H n (F, k) −−−→ H n (L, k) .
For all r, s ≥ 0, we have an exact sequence
(43.1) 0 → H n (L/F, r) → H n (L/F, r + s) → H n (L/F, s).
Proposition 43.2. Let char F 6= 2. Suppose Itn (F ) = 0. Then H n (Fpy /F, k) = 0 for
all k.
Proof. Let α ∈ H n (Fpy /F ). As Fpy is the union of admissible extensions over F (cf.
Definition 31.15), there is an admissible sub-extension L/F of Fpy /F such that αL = 0.
We induct on the degree [L : F√ ] to show αL = 0. Let E be a subfield of L such that
E/F is admissible and L = E( d) with d ∈ D(2h1iE ). It follows from the exactness
of the cohomology sequence (cf. Theorem 99.13) for the quadratic extension L/E that
αE ∈ H n−1 (E) ∪ (d). By Proposition 35.7, the field E satisfies An . Hence all the torsion
Pfister forms hha1 , . . . , an−1 , dii over E are trivial, consequently, H n−1 (E) ∪ (d) = 0 by
Fact 16.2 and therefore αE = 0. By the induction hypothesis, α = 0.
We have shown that H n (Fpy /F ) = 0. Triviality of the group H n (Fpy /F, k) follows
then by induction on k from exactness of the sequence (43.1). ¤
Exercise 43.3. Let char F 6= 2. Show that if H n (Fpy /F ) = 0 then I n (F ) is torsion
free.
Lemma 43.4. A field F of characteristic different from two is pythagorean if and only
if F has no cyclic extensions of degree 4.
Proof. Consider the exact sequence
g b
H 1 (F, 2) −
→ H 1 (F ) −
→ H 2 (F ),
¡ ¢
where b is the Bockstein homomorphism, b (a) = (a) ∪ (−1) (cf. Proposition 101.14).
The field
¡ F¢ is not pythagorean if and only if there is non-square a ∈ F × such that
a ∈ D 2h1i . The later is equivalent to (a) ∪ (−1) = 0 in H 2 (F ) = Br2 (F ) which in its
√
turn is equivalent to (a) ∈ im(g), i.e., the quadratic extension F ( a)/F can be embedded
into a cyclic extension of degree 4. ¤
Let F be a field of characteristic different from two such that µ2n ⊂ F with n > 1 and
m ≤ n. Then Kummer theory implies that the natural map
n m
(43.5) F × /F ×2 = H 1 (F, n) → H 1 (F, m) = F × /F ×2
is surjective.
Lemma 43.6. Let F be a pythagorean field of characteristic different from two. Then
¡ √ ¢
cF (√−1)/F : H 1 F ( −1), s → H 1 (F, s)
is trivial for every s.
1
Proof. If F is non-real then it is quadratically closed,
√ so H (F, s) = 0. Therefore,
we may assume that F is formally real. In particular, F ( −1) 6= F .
Let β ∈ H 1 (F, s + 1) = Hom cont (ΓF , Z/2s+1 Z). Then the kernel of β is an open
subgroup U of ΓF with ΓF /U cyclic of 2-power order. As F is pythagorean, F has no cyclic
extensions of a 2-power order greater than 2 by Lemma 43.4. It follows that [ΓF : U ] ≤ 2
hence β lies in the image of H 1 (F ) → H 1 (F, s + 1). Consequently, β lies in the kernel of
H 1 (F, s + 1) → H 1 (F, s). This shows that the natural map H 1 (F, s + 1) → H 1 (F, s) is
trivial. The statement now follows from the commutativity of the diagram
¡ √ ¢ cF (√−1)/F )
H 1 F ( −1), s + 1 −−−−−−→ H 1 (F, s + 1)
 
 
y y0
¡ √ ¢ √
cF ( −1)/F )
H 1 F ( −1), s −−−−−−→ H 1 (F, s)
¡ √ ¢ ¡ √ ¢
together with the surjectivity of H 1 F ( −1), s + 1 → H 1 F ( −1), s which holds by
√ √
(43.5) as µ2∞ ⊂ Qpy ( −1) ⊂ F ( −1). ¤
√Lemma 43.7. Let F be a field ¡ of ¢characteristic different from two satisfying µ2s ⊂
F ( −1). Then for every d ∈ D 2h1i the class (d) belongs to the image of the natural
map H 1 (Fpy /F, s) → H 1 (Fpy /F ).
1
¡ √ ¢ 1
¡ √ ¢
Proof. By (43.5), the natural
¡ √ map
¢ g : H F ( −1), s → H
¡ √ F ( −1)
¢ is sur-
√ 1
jective. As d ∈ NF ( −1)/F F ( −1) , there exists a γ ∈ H F ( −1), s satisfying
¡ ¢
(d) = g cF (√−1)/F (γ) . By Lemma 43.6, we have cF (√−1)/F (γ) ∈ H 1 (Fpy /F, s) and the
image of cF (√−1)/F (γ) in H 1 (Fpy /F ) coincides with (d). ¤
Theorem 43.8. Let char F 6= 2 and K/F a finite field extension. If I n (K) is torsion
free for some n then I n (F ) is also torsion free.
r
√Proof. Let 2 be the largest power of 2 dividing [K : F ]. Suppose first that the field
F ( −1) contains µ2r+1 .
By Corollary 41.5, we know that I n (F ) is torsion-free if and only F satisfies An .
By Lemma 35.2, it suffices to show¡any ¢bilinear n-fold Pfister form hha1 , . . . , an−1 , dii
with a1 , . . . , an−1 ∈ F × and d ∈ D 2h1i is hyperbolic. By Lemma 43.7, there is an
α ∈ H 1 (Fpy /F, r + 1) such that the natural map H 1 (Fpy /F, r + 1) → H 1 (Fpy /F ) takes α
to (d).
Recall that the graded group H ∗ (Fpy /F, r + 1) has natural structure of a module over
the Milnor ring K∗ (F ) (cf. 101.5). Consider the element
β = {a1 , . . . , an−1 } · α ∈ H n (Fpy /F, r + 1).
As Itn (K) = 0, we have H n (Kpy /K, r + 1) = 0 by Proposition 43.2. Therefore,
[K : F ] · β = cK/F ◦ rK/F (β) = 0
hence 2r β = 0. The composition
H n (F, r + 1) → H n (F ) → H n (F, r + 1)
43. GOING DOWN AND TORSION-FREENESS 193
coincide with the multiplication by 2r . Since the second homomorphism is injective by

(43.1), the image {a1 , . . . , an−1 }·(d) = (a1 , . . . , an−1 , d) of β in H n (F ) is trivial. Therefore,
hha1 , . . . , an−1 , dii is hyperbolic by Fact 16.2.
√
Consider √ the general case. As µ 2 ∞ ⊂ Fpy ( −1) there is a subfield E ⊂ Fpy such that
µ2r+1 ⊂ E( −1) and E/F is an admissible extension. Then L := KE is an admissible
extension of K. In particular, I n (L) is torsion free by Proposition 35.7 and Corollary
41.5. Note also that the degree [L : E] divides [K : F ]. By the first part of the proof
applied to the extension L/E we have Itn (E) = 0. It follows from Theorem 35.12 and
Corollary 41.5 that Itn (F ) = 0. ¤
Corollary 43.9. Let K be a finite extension of a non formally real field F . If
I (K) = 0 then I n (F ) = 0.
n
Proof. If char F = 2, this was shown in Lemma 35.5. If char F 6= 2, this follows
from Theorem 43.8. ¤
Define the torsion-free index of a field F to be
ν(F ) := min{n | I n (F ) is torsion-free}
(or infinity if no such integer exists). Then we have
Corollary 43.10. Let K/F a finitely generated field extension. Then
ν(F ) + tr deg(K/F ) ≤ ν(K).
Proof. If K/F is finite then ν(F ) ≤ ν(K) by Theorem 43.8. It follows from this and
induction that we may assume that K = F (t). If b ∈ Itn (F ) where ν(K) ≤ n + 1 then
hhtii · b = 0 in W (K) as it is torsion in I n+1 (K). It follows by Example 19.13 that b = 0
in W (F ). ¤
Remark 43.11. Let F be a field of characteristic not two.
(1) Define the absolute stability index sta (F ) to be the minimum n (if it exists) such
that I n+1 (F ) = 2I n (F ). Corollary 35.27 implies that
¡ √ ¢
ν F ( −1) ≤ n + 1 if and only if ν(F ) ≤ n + 1 and sta (F ) ≤ n.
(which, by Remark 35.28, is equivalent to I n+1 (L) =¡ 0 for
√ some
¢ quadratic exten-
n+1
sion L/F .) By Fact 16.2 this is equivalent to H F ( −1) = 0.
(2) Let K/F be a finitely generated field ¡ √ extension
¢ of transcendental degree
¡ √ m. In
n+1
[43] it is shown if m > 0 then H F ( −1) = 0 is equivalent to cd2 K( −1)} ≤
n + 1. This is √not true for K/F algebraic, e.g., let F be the quadratic closure of
Q and K = F (3 2).
(3) The “Going Up” Problem of the relationship of ν(F ) and ν(K) even when K/F
is finite is unclear. It is can be shown that ν(K) ≤ [K : F ] ν(F ) when F is
not formally real and that ν(K) ≥ 1 + ν(F ) is possible, but no uniform bound
is known even when F is quadratically closed. Similarly, it can be shown that
sta (K) ≤ [K : F ] sta (F ) but again no uniform bound is known. This compares to
the relative case where Bröcker showed in [21] that the reduced stability satisfied
str (F ) + m ≤ str K ≤ str (F ) + m + 1 using valuation theory.
CHAPTER VIII
On the norm residue homomorphism of degree two
44. The main theorem

In this chapter we prove the degree two case of the Milnor Conjecture (cf. Fact 101.6).
Theorem 44.1. For every field F of characteristic not 2, the norm residue homomor-
phism
hF := h2F : K2 (F )/2K2 (F ) → Br2 (F ),
µ ¶
a, b
taking {a, b} + 2K2 (F ) to the class of the quaternion algebra is an isomorphism.
F
Corollary 44.2. Let F be a field of characteristic not 2. Then
(1) The group Br2 (F ) is generated by the classes of quaternion algebras.
(2) The following is the list of the defining relations between classes of quaternion
algebras:
µ 0 ¶ µ ¶ µ 0 ¶ µ ¶ µ ¶ µ 0¶
aa , b a, b a ,b a, bb0 a, b a, b
(i) = · ; = · ,
µ F ¶ µ F
¶ F F F F
a, b a, b
(ii) · = 1,
µF ¶ F
a, b
(iii) = 1 if a + b = 1.
F
The main idea of the proof is to compare the norm residue homomorphisms hF and
hF (C) , where C is a smooth conic curve over F . The function field F (C) is a generic
splitting field for a symbol in k2 (F ) := K2 (F )/2K2 (F ), so passing from F to F (C) allows
us to carry out inductive arguments.
Theorem 44.1 was originally proven in [102]. The proof used a specialization argument
reducing the problem to the study of the function field of a conic curve and a comparison
theorem of Suslin [133] on the behavior of the norm residue homomorphism over the
function field of a conic curve.
The “elementary” proof presented in this chapter relies neither on a specialization
argument nor on higher K-theory. The key point of the proof is Theorem 46.1. It is also
a consequence of Quillen’s computation of higher K-theory of a conic curve [117, §8, Th.
4.1] and a theorem of Rehmann and Stuhler on the group K2 of a quaternion algebra
given in [118].
Other “elementary” proofs of the bijectivity of hF , avoiding higher K-theory, but still
using a specialization argument were given by Arason in [7] and Wadsworth in [145].
195
196 VIII. ON THE NORM RESIDUE HOMOMORPHISM OF DEGREE TWO
45. Geometry of conic curves

In this section we establish interrelations between projective conic curves and their
corresponding quaternion algebras over a field F of arbitrary characteristic.
45.A. Quaternion algebras and conic curves. Let Q be a quaternion algebra
over a field F . Recall (cf. Appendix 98.E) that Q carries a canonical involution a 7→ ā,
the reduced trace linear map
Trd : Q → F, a 7→ a + ā
and the reduced norm quadratic map
Nrd : Q → F, a 7→ aā.
Every element a ∈ Q satisfies the quadratic equation
a2 − Trd(a)a + Nrd(a) = 0.
Set
VQ := ker(Trd) = {a ∈ Q | ā = −a},
so VQ is a 3-dimensional subspace of Q. Note that x2 = − Nrd(x) ∈ F for any x ∈ VQ ,
and the map ϕQ : VQ → F given by ϕQ (x) = x2 is a quadratic form on VQ . The space VQ
is the orthogonal complement to 1 in Q with respect to the non-degenerate bilinear form
on Q:
(a, b) 7→ Trd(ab).
The quadric CQ of the form ϕQ (x) in the projective plane P(VQ ) is a smooth projective
conic curve. In fact, every smooth projective conic curve (1-dimensional quadric) is of
the form CQ for some quaternion algebra Q (cf. Exercise 12.6).
Proposition 45.1. The following conditions are equivalent:
(1) Q is split.
(2) CQ is isomorphic to the projective line P1 .
(3) CQ has a rational point.
Proof. (1) ⇒ (2): The algebra Q is isomorphic to the matrix algebra M2 (F ). Hence
VQ is the space of trace 0 matrices and CQ is given by the equation X 2 + Y Z = 0. The
morphism CQ → P1 , given by [X : Y : Z] 7→ [X : Y ] = [−Z : X] is an isomorphism.
(2) ⇒ (3) is obvious.
(3) ⇒ (1): There is a nonzero element x ∈ Q such that x2 = 0. In particular, Q is not
a division algebra and therefore Q is split. ¤
Corollary 45.2. Let Q be a quaternion algebra. Then
(1) Every divisor on CQ of degree zero is principal.
(2) If Q is a division algebra, the degree of every closed point of CQ is even.
Proof. (1): Let w be a divisor on CQ of degree zero. Let K/F be a separable qua-
dratic splitting field of Q and σ the nontrivial automorphism of K over F . By Proposition
45.1, the conic CQ is isomorphic to the projective line over K. As deg σ(wK ) = deg wK ,
there is a function f ∈ K(CQ )× with div(f ) = wK . Let s = f · σ(f −1 ). Since div(s) =
wK − σ(wK ) = 0 we have s ∈ K × and NK/F (s) = s · σ(s) = 1. By the Hilbert theorem
45. GEOMETRY OF CONIC CURVES 197
90, there is t ∈ K × with s = σ(t) · t−1 . Setting g = tf we have σ(g) = g, i.e., g ∈ F (CQ )×
and div(g) = w.
(2): Let x ∈ CQ be a closed point. As CQ has a rational point over F (x), by Proposition
45.1, the algebra Q is split over F (x). It follows from Corollary 98.6 that the degree
[F (x) : F ] is even. ¤
Example 45.3. If char F 6= 2, there is a basis 1, i, j, k of Q such that a := i2 ∈ F × ,
b := j 2 ∈ F × , k = ij = −ji (cf. Example 98.11). Then VQ = F i ⊕ F j ⊕ F k and CQ is
given by the equation aX 2 + bY 2 − abZ 2 = 0.
Example 45.4. If char F = 2, there is a basis 1, i, j, k of Q such that a := i2 ∈ F ,
b := j 2 ∈ F , k = ij = ji + 1 (cf. Example 98.12). Then VQ = F 1 ⊕ F i ⊕ F j and CQ is
given by the equation X 2 + aY 2 + bZ 2 + Y Z = 0.
For every a ∈ Q define the F -linear function la on VQ by the formula
la (x) = Trd(ax).
The reduced trace form Trd is a non-degenerate bilinear form on Q. Indeed to see this
this is sufficient to check over a splitting field of Q where Q is isomorphic to a matrix
algebra, and it is clear in this case. Hence every F -linear function on VQ is equal to la for
some a ∈ Q.
Lemma 45.5. Let a, b ∈ Q and α, β ∈ F . Then

(1) la = lb if and only if a − b ∈ F .
(2) lαa+βb = αla + βlb .
(3) la = −la .
(4) la−1 = −(Nrd a)−1 · la if a is invertible.
Proof. (1): This follows from the fact that VQ is orthogonal to F with respect to
the bilinear form Trd.
(2) is obvious.
(3): For any x ∈ VQ , we have
la (x) = Trd(āx) = Trd(āx) = Trd(x̄a) = − Trd(xa) = − Trd(ax) = −la (x).
(4): It follows from (2) and (3) that (Nrd a)la−1 = la = −la . ¤
Every element a ∈ Q \ F generates a quadratic subalgebra F [a] = F ⊕ F a of Q.
Conversely, every quadratic subalgebra K of Q is of the form F [a] for any a ∈ K \ F . By
Lemma 45.5, the linear form la on VQ is independent, up to a multiple in F × , of the choice
of a ∈ K \ F . Hence the line in P(VQ ) given by the equation la (x) = 0 is determined by
K. The intersection of this line with the conic CQ is an effective divisor on CQ of degree
two. Thus, we have the following maps
Quadratic Rational points Lines Degree 2 effective

→ = →
subalgebras of Q in P(VQ∗ ) in P(VQ ) divisors on CQ
Proposition 45.6. The two maps above are bijections.

Proof. The first map is a bijection since every line in P(VQ ) is given by an equation
la = 0 for some a ∈ Q \ F and a generates a quadratic subalgebra of Q. The last map
is a bijection since the embedding of CQ as a closed subscheme of P(VQ ) is given by a
complete linear system. ¤
Remark 45.7. Effective divisors on CQ of degree 2 are rational points of the symmetric
square S 2 (CQ ). Proposition 45.6 essentially asserts that S 2 (CQ ) is isomorphic to the
projective plane P(VQ∗ ).
Suppose Q is a division algebra. Then the conic curve CQ has no rational points and
quadratic subalgebras of Q are quadratic (maximal) subfields of Q. An effective cycle on
CQ of degree 2 is a closed point of degree 2. Thus, by Proposition 45.6, we have bijections
Quadratic ∼ Rational points Lines ∼ Points of

→ = →
subfields of Q in P(VQ )
∗
in P(VQ ) degree 2 in CQ
In what follows, we shall frequently use the bijection constructed above between the
set of quadratic subfields of Q and the set of degree 2 closed points of CQ .
45.B. Key identity. In the following proposition, we write a multiple of the qua-
dratic form ϕQ on VQ as a degree two polynomial of linear forms.
Proposition 45.8. Let Q be a quaternion algebra over F . For any a, b, c ∈ Q,

¡ ¢
lab · lc + lbc · la + lca · lb = Trd(cba) − Trd(abc) · ϕQ .
Proof. We write T for Trd in this proof. For every x ∈ VQ we have:
lab (x) · lc (x) = T (abx)T (cx)

¡ ¢
= T a(T (b) − b)x T (cx)
= T (ax)T (b)T (cx) − T (abx)T (cx)
¡ ¢
= T (ax)T (b)T (cx) − T abT (cx)x
= T (ax)T (b)T (cx) − T (abc)x2 + T (abxcx),
lbc (x) · la (x) = T (bcx)T (ax)

¡ ¢
= T (T (b) − b)cx T (ax)
= T (cx)T (b)T (ax) − T (bcx)T (ax)
¡ ¢
= −T (ax)T (b)T (cx) − T bcx(ax + x̄ā)
= −T (ax)T (b)T (cx) − T (bcxax) + T (bca)x2
= −T (ax)T (b)T (cx) − T (axbcx) + T (cba)x2
45. GEOMETRY OF CONIC CURVES 199
lca (x) · lb (x) = T (cax)T (bx)

= −T (acx)T (bx)
¡ ¢
= −T aT (bx)cx
= −T (abxcx) + T (axbcx).
Adding the equalities yields the result. ¤

45.C. Residue fields of points of CQ and quadratic subfields of Q. Suppose
that the quaternion algebra Q is a division algebra. Recall that quadratic subfields of Q
correspond bijectively to degree 2 points of CQ . We shall show how to identify a quadratic
subfield of Q with the residue field of the corresponding point in CQ of degree 2.
Choose a quadratic subfield K ⊂ Q. For every a ∈ Q \ K, one has Q = K ⊕ aK. We
define a map
µa : VQ∗ → K
by the rule: if c = u + av for u, v ∈ K then µa (lc ) = v. Clearly,
µa (lc ) = 0 if and only if c ∈ K.
By Lemma 45.5, the map µa is well defined and F -linear. If b ∈ Q \ K is another element,
we have
(45.9) µb (lc ) = µb (la )µa (lc ),
hence the maps µa and µb differ by the multiple µb (la ) ∈ K × . The map µa extends to an
F -algebra homomorphism
µa : S • (VQ∗ ) → K
in the usual way (where S • denotes the symmetric algebra).
Let x ∈ CQ ⊂ P(VQ ) be the point of degree 2 corresponding to the quadratic subfield
K. The local ring OP(VQ ),x is the subring of the quotient field of the symmetric algebra
S • (VQ∗ ) generated by the fractions lc /ld for all c ∈ Q and d ∈ Q \ K.
Fix an element a ∈ Q \ F . We define an F -algebra homomorphism
µ : OP(VQ ),x → K
by the formula
µ(lc /ld ) = µa (lc )/µa (ld ).
Note that µa (ld ) 6= 0 since d ∈
/ K and the map µ is independent of the choice of a ∈ Q \ K
by (45.9).
We claim that the map µ vanishes on the quadratic form ϕQ defining CQ in P(VQ ).
Proposition 45.8 gives a formula for a multiple of the quadratic form ϕQ with the coefficient
α := Trd(cba) − Trd(abc).
Lemma 45.10. There exist a ∈ Q \ K, b ∈ K, and c ∈ Q such that α 6= 0.
Proof. Pick¡ any b ∈ K
¢ \ F and any a ∈ Q such that ab 6= ba. Clearly, a ∈ Q \ K.
Then α = Trd (ba − ab)c is nonzero for some c ∈ Q since the bilinear form Trd is
non-degenerate on Q. ¤
Choose a, b and c as in Lemma 45.10. We have µa (lb ) = 0 since b ∈ K. Also µa (la ) = 1

and µa (lab ) = b̄. Write c = u + av for u, v ∈ K then µa (lc ) = v. As
bc̄ = bū + bv̄ā = bū + Trd(bv̄ā) − av b̄,
we have µa (lbc ) = −v b̄ and by Proposition 45.8,
αµ(ϕQ ) = µa (lab )µa (lc ) + µa (lbc )µa (la ) + µa (lca )µa (lb ) = b̄v − v b̄ = 0.
Since Q is a division algebra and α 6= 0, we have µ(ϕQ ) = 0 as claimed.

The local ring OCQ ,x coincides with the factor ring OP(VQ ),x /ϕQ OP(VQ ),x . Therefore, µ
factors through an F -algebra homomorphism
µ : OCQ ,x → K.
Let e ∈ K \ F . The function le /la is a local parameter of the local ring OCQ ,x , i.e., it
generates the maximal ideal of OCQ ,x . Since µ(le /la ) = 0, the map µ induces a field
isomorphism
∼
(45.11) F (x) → K
of degree 2 field extensions of F . We have proved
Proposition 45.12. Let Q be a division quaternion algebra. Let K ⊂ Q be a quadratic

subfield and x ∈ CQ the corresponding point of degree 2. Then the residue field F (x) is
canonically isomorphic to K over F . Let a ∈ Q and b ∈ Q \ K. Write a = u + bv for
unique u, v ∈ K. Then the value (la /lb )(x) in F (x) of the function la /lb at the point x
corresponds to the element v ∈ K under the isomorphism (45.11).
46. Key exact sequence
¡In this¢ section we prove exactness of a sequence comparing the groups K2 (F ) and
K2 F (C) over a field F of arbitrary characteristic.
Let C be a smooth curve over a field F . For every (closed) point x ∈ C, there is a
residue homomorphism
¡ ¢ ¡ ¢
∂x : K2 F (C) → K1 F (x) = F (x)×
induced by the discrete valuation of the local ring OC,x (cf. (49.A)).
In this section we prove the following:
Theorem 46.1. Let C be a conic curve over a field F . Then the sequence
¡ ¢ ∂ a c
K2 (F ) → K2 F (C) − → F (x)× − → F ×,
x∈C
with ∂ = (∂x ) and c = (NF (x)/F ), is exact.

46. KEY EXACT SEQUENCE 201
¡ ¢
46.A. Filtration on K2 F (C) . Let C be a conic over F . If C has a¡rational ¢ point,
i.e., C ' P1F , the statement of Theorem 46.1 is Milnor’s computation of K2 F (t) given in
Theorem 100.7. So we may (and will) assume that C has no rational point. By Corollary
45.2, the degree of every closed point of C is even.
Fix a closed point x0 ∈ C of degree 2. As in §29, for any n ∈ Z let Ln be the
F -subspace © ª © ª
f ∈ F (C)× | div(f ) + nx0 ≥ 0 ∪ 0
of F (C). Clearly Ln = 0 if n < 0. Recall that L0 = F and Ln · Lm ⊂ Ln+m . It follows
from Lemma 29.7 that dim Ln = 2n + 1 if n ≥ 0.
We write L× ×
n for Ln \ {0}. Note that the value g(x) in F (x) is defined for every g ∈ Ln
and point x 6= x0 .
Since any divisor on C of degree zero is principal by Corollary 45.2, for every point
x ∈ C of degree 2n we can choose a function px ∈ L× n satisfying div(px ) = x − nx0 .
In particular, px0 ∈ F × . Note that px is uniquely determined up to a scalar multiple.
Clearly, px (x) = 0 if x 6= x0 . Every function in L× n can be written as the product of a
nonzero constant and finitely many px for some points x of degree at most 2n.
Lemma 46.2. Let x ∈ C be a point of degree 2n different from x0 . If g ∈ Lm satisfies
g(x) = 0 then g = px q for some q ∈ Lm−n . In particular, g = 0 if m < n.
Proof. Consider the F -linear map
ex : Lm → F (x), ex (g) = g(x).
If m < n, the map ex is injective since x does not belong to the support of the divisor of
a function in L×m . Suppose that m = n and g ∈ ker ex . Then div(g) = x − nx0 and hence
g is a multiple of px . Thus, the kernel of ex is 1-dimensional. By dimension count, ex is
surjective.
Therefore, for arbitrary m ≥ n, the map ex is surjective and
dim ker(ex ) = dim Lm − deg(x) = 2m + 1 − 2n.
The image of the injective linear map Lm−n → Lm given by multiplication by px is
contained in ker(ex ) and of dimension dim Lm−n = 2m + 1 − 2n. Therefore, ker(ex ) =
px Lm−n . ¤
¡ ¢
For every n ∈ Z, let Mn be the subgroup of K2 F (C) generated by the symbols {f, g}
with f, g ∈ L× × ×
n , i.e., Mn = {Ln , Ln }. We have the following filtration:
¡ ¢
(46.3) 0 = M−1 ⊂ M0 ⊂ M1 ⊂ · · · ⊂ K2 F (C) .
¡ ¢
Note
¡ that
¢ M0 coincides with the image of the homomorphism K 2 (F ) → K 2 F (C) and
K2 F (C) is the union of all the Mn as the group F (C)× is the union of the subsets L× n.
×
If f ∈ Ln , the degree of every point in the support of div(f ) is at most 2n. In
particular, ∂x (Mn−1 ) = 0 for every point x of degree 2n. Therefore, for every n ≥ 0, we
have a well defined homomorphism
a
∂n : Mn /Mn−1 → F (x)×
deg x=2n
induced by ∂x over all points x ∈ C of degree 2n.

We refine the filtration (46.3) by adding an extra term M 0 between M0 and M1 . Set
M 0 := {L× × × × 0
1 , L0 } = {L1 , F }, so the group M is generated by M0 and symbols of the
form {px , α} for all points x ∈ C of
` degree 2 and all α ∈ F × .
Denote by A0 the subgroup of deg x=2 F (x)× consisting of all families (αx ) such that
Q
αx ∈ F × for all x and x αx = 1. Clearly, ∂1 (M 0 /M0 ) ⊂ A0 .
Theorem 46.1 is a consequence of the following three propositions.
Proposition 46.4. If n ≥ 2, the map
a
∂n : Mn /Mn−1 → F (x)×
deg x=2n
is an isomorphism.
Proposition 46.5. The restriction ∂ 0 : M 0 /M0 → A0 of ∂1 is an isomorphism.
Proposition 46.6. The sequence
Ã !
a
0 ∂1 c
0 → M1 /M −
→ F (x)× /A0 −
→ F×
deg x=2
is exact.
¡ ¢
Proof of Theorem 46.1. Since K2 F (C) is the union of Mn , it is sufficient to prove
that the sequence
∂
a c
0 → Mn /M0 →− F (x)× −
→ F×
deg x≤2n
is exact for every n ≥ 1. We induct on n. The case n = 1 follows from Propositions 46.5
and 46.6. The induction step is guaranteed by Proposition 46.4. ¤
46.B. Proof of Proposition 46.4. To effect the proof, we shall construct the inverse
map of ∂n . We need the following two lemmas.
Lemma 46.7. Let x ∈ C be a point of degree 2n > 2. Then for every u ∈ F (x)× , there
exist f ∈ L× ×
n−1 and h ∈ L1 such that (f /h)(x) = u.
Proof. The F -linear map

ex : Ln−1 → F (x), f 7→ f (x)
is injective by Lemma 46.2. Hence
dim coker(ex ) = deg(x) − dim Ln−1 = 2n − (2n − 1) = 1.
Consider the F -linear map
g : L1 → coker(ex ), g(h) = u · h(x) + im(ex ).
Since dim L1 = 3, the kernel of g contains a nonzero function h ∈ L× 1. We have
u · h(x) = f (x) for some f ∈ L×
n−1 . Since deg x > 2 the value h(x) is nonzero. Hence
u = (f /h)(x). ¤
Let x ∈ C be a point of degree 2n > 2. Define a map

ψx : F (x)× → Mn /Mn−1
as follows: By Lemma 46.7, for each element u ∈ F (x)× , we can choose f ∈ L×
n−1 and
×
h ∈ L1 such that (f /h)(x) = u. We set
n fo
ψx (u) = px , + Mn−1 .
h
Lemma 46.8. The map ψx is a well-defined homomorphism.
Proof. Let f 0 ∈ L× 0 × 0 0
n−1 and h ∈ L1 be two functions with (f /h )(x) = u. Then
f 0 h − f h0 ∈ Ln and (f 0 h − f h0 )(x) = 0. By Lemma 46.2, we have f 0 h − f h0 = λpx for
some λ ∈ F . If λ = 0, then f /h = f 0 /h0 .
Suppose λ 6= 0. Since (λpx )/(f 0 h) + (f h0 )/(f 0 h) = 1, we have
n λp f h0 o n f o n f 0 o
x
0= , ≡ px , − px , 0 mod Mn−1 .
f h f 0h
0 h h
Hence
{px , f /h} + Mn−1 = {px , f 0 /h0 } + Mn−1 ,
so that the map ψ is well defined.
Let u3 = u1 u2 ∈ F (x)× . Choose fi ∈ L× ×
n−1 and hi ∈ L1 satisfying (fi /hi )(x) = ui for
i = 1, 2, 3. The function f1 f2 h3 − f3 h1 h2 belongs to L2n−1 and has zero value at x. We
have f1 f2 h3 − f3 h1 h2 = px q for some q ∈ Ln−1 by Lemma 46.2. Since (px q)/(f1 f2 h3 ) +
(f3 h1 h2 )/(f1 f2 h3 ) = 1,
n p q fhh o n f3 o n f1 o n f2 o
x 3 1 2
0= , ≡ px , − px , − px , mod Mn−1 .
f1 f2 h3 f1 f2 h3 h3 h1 h2
Thus, ψx (u3 ) = ψx (u1 ) + ψx (u2 ). ¤
By Lemma 46.8, we have a homomorphism
X a
ψn = ψx : F (x)× → Mn /Mn−1 .
deg x=2n
We claim that ∂n and ψn are isomorphisms inverse to each other. If x is a point of degree
2n > 2 and u ∈ F (x)× , choose f ∈ L× ×
n−1 and h ∈ L1 such that (f /h)(x) = u. We have
¡n f o¢ ³ f ´
∂x px , = (x) = u
h h
and the symbol {px , f /h} has no nontrivial residues at other points of degree 2n. There-
fore, ∂n ◦ ψn is the identity.
To finish the proof of Proposition 46.4, it suffices to show that ψn is surjective. The
group Mn /Mn−1 is generated by classes of the form {px , g} + Mn−1 and {px , py } + Mn−1 ,
where g ∈ L× n−1 and x, y are distinct points of degree 2n. Clearly
¡ ¢
{px , g} + Mn−1 = ψx g(x) ,
hence {px , g} + Mn−1 ∈ im ψn .
By Lemma 46.7, there are elements f ∈ L× ×

n−1 and h ∈ L1 such that px (y) = (f /h)(y).
The function px h − f belongs to L×n+1 and has zero value at y. Therefore px h − f = py q
×
for some q ∈ L1 by Lemma 46.2. Since (py q)/(px h) + f /(px h) = 1 we have
np q f o
y
0= , ≡ {px , py } mod im(ψn ). ¤
px h px h
46.C. Proof of Proposition 46.5. Define a homomorphism
ρ : A0 → M 0 /M0
by the rule ³a ´ X
ρ αx = {px , αx } + M0 .
deg x=2
−1
Since ∂x {px , α} = α and ∂x0 {px , α} = α for every x 6= x0 and the product of all αx is
equal to 1, the composition ∂ 0 ◦ ρ is the identity. Clearly, ρ is surjective. ¤
46.D. Generators and relations of A(Q)/A0 . It remains to prove Proposition 46.6.
∼
Let Q be a quaternion division algebra satisfying C → CQ . By Proposition 45.12, the
norm homomorphism a
F (x)× → F ×
deg x=2
is canonically isomorphic to the norm homomorphism
a
(46.9) K × → F ×,
where the coproduct is taken over all quadratic subfields K ⊂ Q. Note that the norm
map NK/F : K × → F × is the restriction of the reduced norm Nrd on K. Let A(Q) be
the kernel of the norm
` homomorphism (46.9). Under the above canonical isomorphism
0 ×
the subgroup A of F (x) corresponds to the subgroup of A(Q) Q (that we still 0denote
by A0 ) consisting of all families
` × K (a ) satisfying aK ∈ F ×
and aK = 1, i.e., A is the
intersection of A(Q) and F . Therefore Proposition 46.6 asserts that the canonical
homomorphism
(46.10) ∂1 : M1 /M 0 → A(Q)/A0
is an isomorphism. In the proof of Proposition 46.6, we shall construct the inverse iso-
morphism. In order to do so, it is convenient to have a presentation of the group A(Q)/A0
by generators and relations.
We define a map (not a homomorphism!)
³a ´
Q× → K × /A0 , a 7→ ea
as follows: If a ∈ Q× is not a scalar, it is contained in a ùnique quadratic subfield K
of Q. Therefore, a defines
¡` an ¢ element of the coproduct K × . We denote by e a the
corresponding class in K × /A0 . If a ∈ F × , of course, a belongs to all quadratic
¡` × ¢ 0
subfields. Nevertheless a defines a unique element ea of the factor group K /A (we
place a in any quadratic subfield). Clearly
(46.11) g =e
(ab) a · eb if a and b commute.
(Note that we are using ` multiplicative

¢ notation for the operation in the factor group.)
Obviously, the group ( K × /A0 is an abelian group generated by the e a for all a ∈ Q×
with the set of defining relations given by (46.11).
The group A(Q)/A0 is generated (as an abelian group) by the products e a1 e
a2 · · · e
an with
×
ai ∈ Q and Nrd(a1 a2 · · · an ) = 1, with the following set of defining relations:
(1) (e
a1ea2 · · · e
an ) · (e
an+1e
an+2 · · · e
an+m ) = (e a1 e
a2 · · · e
an+m ).
(2) e ]
aeb(a−1 )(b] −1 ) = 1.
(3) If ai−1 and ai commute, then e a1 · · · e

ai−1eai · · · e
an = e
a1 · · · a^
i−1 ai · · · e
an .
The set of generators is too large for our purposes. In the next subsection, we shall
find another presentation of A(Q)/A0 (cf. Corollary 46.27). More precisely, we shall define
an abstract group G by generators and relations (with a “better” set of generators) and
prove that G is isomorphic to A(Q)/A0 .
46.E. The group G. Let Q be a division quaternion algebra over a field F . Consider
the abelian group G defined by generators and relations as follows: The sign ∗ will denote
the operation in G with 1 its identity element.
Generators: Symbols (a, b, c) for all ordered triples with a, b, c elements of Q× satis-
fying abc = 1. Note that if (a, b, c) is a generator of G then so are the cyclic permutations
(b, c, a) and (c, a, b).
Relations: ¡
(R1) : (a, b, cd) ∗ (ab, c, d) = (b, c, da) ∗ bc, d, a) for all a, b, c, d ∈ Q× satisfying abcd = 1;
(R2) : (a, b, c) = 1 if a and b commute.
For an (ordered) sequence a1 , a2 , . . . , an (n ≥ 1) of elements in Q× satisfying a1 a2 . . . an =

1, we define a symbol
(a1 , a2 , . . . , an ) ∈ G
by induction on n as follows. The symbol is trivial if n = 1 or 2. If n ≥ 3, we set
(a1 , a2 , . . . , an ) := (a1 , a2 , . . . , an−2 , an−1 an ) ∗ (a1 a2 · · · an−2 , an−1 , an ).
Note that if a1 a2 . . . an = 1 then a2 . . . an a1 = 1.
Lemma 46.12. The symbols do not change under cyclic permutations, i.e.,
(a1 , a2 , . . . , an ) = (a2 , . . . , an , a1 ) if a1 a2 . . . an = 1.
Proof. We induct on n. The statement is clear if n = 1 or 2. If n = 3,
(a1 , a2 , a3 ) = (a1 , a2 , a3 ) ∗ (a1 a2 , a3 , 1) (relation R2)

= (a2 , a3 , a1 ) ∗ (a2 a3 , 1, a1 ) (relation R1)
= (a2 , a3 , a1 ) (relation R2).
Suppose that n ≥ 4. We have

(a1 , a2 , . . . , an ) =(a1 , . . . , an−2 , an−1 an ) ∗ (a1 a2 · · · an−2 , an−1 , an ) (definition)
=(a2 , . . . , an−2 , an−1 an , a1 ) ∗ (a1 a2 · · · an−2 , an−1 , an ) (induction)
=(a2 , . . . , an−2 , an−1 an a1 ) ∗ (a2 a3 · · · an−2 , an−1 an , a1 )
∗ (a1 a2 · · · an−2 , an−1 , an ) (definition)
=(a2 , . . . , an−2 , an−1 an a1 ) ∗ (a1 , a2 a3 · · · an−2 , an−1 an )
∗ (a1 a2 · · · an−2 , an−1 , an ) (case n = 3)
=(a2 , . . . , an−2 , an−1 an a1 ) ∗ (a2 a3 · · · an−2 , an−1 , an a1 )
∗ (a2 a3 · · · an−1 , an , a1 ) (relation R1)
=(a2 , . . . , an−2 , an−1 , an a1 ) ∗ (a2 a3 · · · an−1 , an , a1 ) (definition)
=(a2 , . . . , an , a1 ) (definition). ¤
Lemma 46.13. If a1 a2 . . . an = 1 and ai−1 commutes with ai for some i, then
(a1 , . . . , ai−1 , ai , . . . , an ) = (a1 , . . . , ai−1 ai , . . . , an ).
Proof. We may assume that n ≥ 3 and i = n by Lemma 46.12. We have
(a1 , . . . , an−2 , an−1 , an ) = (a1 , . . . , an−2 , an−1 an ) ∗ (a1 a2 · · · an−2 , an−1 , an ) (definition)
= (a1 , . . . , an−2 , an−1 an ) (relation R2). ¤
Lemma 46.14. (a1 , . . . , an ) ∗ (b1 , . . . , bm ) = (a1 , . . . , an , b1 , . . . , bm ).
Proof. We induct on m. By Lemma 46.13, we may assume that m ≥ 3. We have
L.H.S. = (a1 , . . . , an ) ∗ (b1 , . . . , bm−1 bm ) ∗ (b1 b2 · · · bm−2 , bm−1 , bm ) (definition)
= (a1 , . . . , an , b1 , . . . , bm−1 bm ) ∗ (b1 b2 · · · bm−2 , bm−1 , bm ) (induction)
= (a1 , . . . , an , b1 , . . . , bm ) (definition). ¤
As usual, we write [a, b] for the commutator aba−1 b−1 .
Lemma 46.15. Let a, b ∈ Q× .
(1) For every nonzero b0 ∈ F b + F ba, one has [a, b] = [a, b0 ]. Similarly, [a, b] = [a0 , b]
for every nonzero a0 ∈ F a + F ab.
(2) For every nonzero b0 ∈ F b + F ba + F bab there exists a0 ∈ Q× such that [a, b] =
[a , b] = [a0 , b0 ].
0
Proof. (1): We have b0 = bx, where x ∈ F + F a. Hence x commutes with a so

[a, b] = [a, b0 ]. The proof of the second statement is similar.
(2): There is nonzero a0 ∈ F a + F ab such that b0 ∈ F b + F ba0 . By the first part,
[a, b] = [a0 , b] = [a0 , b0 ]. ¤
Corollary 46.16. (1) Let [a, b] = [c, d]. Then there are a0 , b0 ∈ Q× such that [a, b] =
[a0 , b] = [a0 , b0 ] = [c, b0 ] = [c, d].
(2) Every pair of commutators in Q× can be written in the form [a, b] and [c, d] with
b = c.
Proof. (1): If [a, b] = 1 = [c, d], we can take a0 = b0 = 1. Otherwise, both sets
{b, ba, bab} and {d, dc} are linearly independent. Let b0 be a nonzero element in the
intersection of the subspaces F b + F ba + F bab and F d + F dc. The statement follows from
Lemma 46.15.
(2): Let [a, b] and [c, d] be two commutators. We may clearly assume that [a, b] 6= 1 6=
[c, d], so that both sets {b, ba, bab} and {c, cd} are linearly independent. Choose a nonzero
element b0 in the intersection of F b + F ba + F bab and F c + F cd. By Lemma 46.15, we
have [a, b] = [a0 , b0 ] for some a0 ∈ Q× and [c, d] = [b0 , d]. ¤
Lemma 46.17. Let h ∈ Q× . The following conditions are equivalent:

(1) h = [a, b] for some a, b ∈ Q× .
(2) h ∈ [Q× , Q× ].
(3) Nrd(h) = 1.
Proof. The implications (1) ⇒ (2) ⇒ (3) are obvious.

(3) ⇒ (1): Let K be a separable quadratic subfield containing h. (If h is purely insepa-
rable, then h2 ∈ F and therefore h = 1.) Since NK/F (h) = Nrd(h) = 1, by the Hilbert
theorem 90, we have h = b̄b−1 for some b ∈ K × . By the Noether-Skolem Theorem,
b̄ = aba−1 for some a ∈ Q× , hence h = [a, b]. ¤
Let h ∈ Q× satisfy Nrd(h) = 1. Then h = [a, b] = aba−1 b−1 for some a, b ∈ Q× by

Lemma 46.17. Consider the following element
b
h = (b, a, b−1 , a−1 , h) ∈ G.
Lemma 46.18. The element b

h does not depend on the choice of a and b.
Proof. Let h = [a, b] = [c, d]. By Corollary 46.16(1), we may assume that either
a = c or b = d. Consider the first case (the latter case is similar). We can write d = bx,
where x commutes with a. We have
(d, a, d−1 , a−1 , h) = (bx, a, x−1 b−1 , a−1 , h)

= (bx, x−1 , b−1 ) ∗ (b, x, a, x−1 b−1 , a−1 , h) (Lemmas 46.13, 46.14)
= (bx, x−1 , b−1 ) ∗ (a−1 , h, b, x, a, x−1 b−1 ) (Lemma 46.12)
= (a−1 , h, b, x, a, x−1 b−1 , bx, x−1 , b−1 ) (Lemma 46.14)
= (a−1 , h, b, a, b−1 ) (Lemma 46.13)
= (b, a, b−1 , a−1 , h) (Lemma 46.12). ¤
Lemma 46.19. For every h1 , h2 ∈ [Q× , Q× ] we have
hd b b −1 −1
1 h2 = h1 ∗ h2 ∗ (h1 h2 , h2 , h1 ).
Proof. By Corollary 46.16(2), we have h1 = [a1 , c] and h2 = [c, b2 ] for some a1 , b2 , c ∈

Q× . Then h1 h2 = [a1 b−1 −1
2 , b2 cb2 ] and
hb1 ∗ hb2 ∗ (h1 h2 , h−1 , h−1 ) = (c, a1 , c−1 , a−1 , h1 , h2 , b2 , c, b−1 , c−1 ) ∗ (h1 h2 , h−1 , h−1 )
2 1 1 2 2 1
−1 −1 −1 −1 −1 −1
= (b2 , c, b2 , c , c, a1 , c , a1 , h1 , h2 ) ∗ (h2 , h1 h1 h2 )
= (b2 , c, b−1 −1 −1
2 , a1 , c , a 1 , h 1 h 2 )
= (b2 , c, b−1 −1 −1 −1 −1 −1
2 , b2 c b2 ) ∗ (b2 cb2 , a1 , c , a1 , h1 h2 )
= (b−1 −1 −1 −1 −1 −1
2 , b2 c b2 , b2 , c) ∗ (c , a1 , h1 h2 , b2 cb2 , a1 )
= (b−1 −1 −1 −1 −1 −1
2 , b2 c b2 , b2 , c, c , a1 , h1 h2 , b2 cb2 , a1 )
= (b−1 −1 −1 −1 −1 −1 −1
2 , b2 c b2 , b2 , a1 , h1 h2 , b2 cb2 , a1 b2 , b2 a1 , a1 )
= (b2 , a−1 −1 −1 −1 −1 −1 −1
1 , h1 h2 , b2 cb2 , a1 b2 , b2 c b2 ) ∗ (b2 a1 , a1 , b2 )
= (b2 a−1 −1 −1 −1 −1 −1 −1
1 , a1 , b2 , b2 , a1 , h1 h2 , b2 cb2 , a1 b2 , b2 c b2 )
= (b2 a−1 −1 −1 −1 −1
1 , h1 h2 , b2 cb2 , a1 b2 , b2 c b2 )
= (b2 cb−1 −1 −1 −1 −1
2 , a1 b2 , b2 c b2 , b2 a1 , h1 h2 )
= hd
1 h2 . ¤
Let a1 , a2 , . . . , an ∈ Q× satisfying Nrd(h) = 1 where h = a1 a2 . . . an . We set
(46.20) ((a1 , a2 , . . . , an )) := (a1 , a2 , . . . , an , h−1 ) ∗ b
h ∈ G.
Lemma 46.21. ((a1 , a2 , . . . , an )) ∗ ((b1 , b2 , . . . , bm )) = ((a1 , . . . , an , b1 , . . . , bm )).
Proof. Set h := a1 · · · an and h0 := b1 · · · bm . We have
L.H.S. = (a1 , a2 , . . . , an , h−1 ) ∗ (b1 , b2 , . . . , bm , (h0 )−1 ) ∗ b
h ∗ hb0
= (a1 , a2 , . . . , an , b1 , b2 , . . . , bm , (h0 )−1 , h−1 ) ∗ b
h ∗ hb0
= (a1 , a2 , . . . , an , b1 , b2 , . . . , bm , (hh0 )−1 ) ∗ (hh0 , (h0 )−1 , h) ∗ b
h ∗ hb0
c0 (Lemma 46.19)
= (a1 , a2 , . . . , an , b1 , b2 , . . . , bm , (hh0 )−1 ) ∗ hh
= R.H.S. ¤
The following Lemma is a consequence of the definition (46.20) and Lemma 46.13.
Lemma 46.22. If ai−1 commutes with ai for some i, then
((a1 , . . . , ai−1 , ai , . . . , an )) = ((a1 , . . . , ai−1 ai , . . . , an )).
Lemma 46.23. ((a, b, a−1 , b−1 )) = 1.
Proof. Set h = [a, b]. We have
L.H.S. = (a, b, a−1 , b−1 , h−1 ) ∗ b
h = (a, b, a−1 , b−1 , h−1 ) ∗ (b, a, b−1 , a−1 , h) = 1. ¤
We want to establish an isomorphism between G and A(Q)/A0 . To do so we define a
map π : G → A(Q)/A0 by the formula
¡ ¢
π (a, b, c) = ãb̃c̃ ∈ A(Q)/A0 ,
where a, b, c ∈ Q× satisfy abc = 1. Clearly, π is well defined.

Let a1 , a2 , . . . , an ∈ Q× with a1 a2 · · · an = 1. By induction on n, we have
¡ ¢
π (a1 , a2 , . . . , an ) = e
a1e an ∈ A(Q)/A0 .
a2 · · · e
Hence π is a homomorphism by Lemma 46.14.
Let h ∈ [Q× , Q× ]. Write h = [a, b] for a, b ∈ Q× . We have
¡ ¢
π(b
h) = π (b, a, b−1 , a−1 , h) = e
h.
If a1 , a2 , . . . , an ∈ Q× satisfies Nrd(h) = 1 with h = a1 a2 · · · an , then
¡ ¢ ¡ ¢
(46.24) π ((a1 , a2 , . . . , an )) = π (a1 , a2 , . . . , an , h−1 ) ∗ π(b
h) = e
a1e
a2 · · · e
an .
Define a homomorphism θ : A(Q)/A0 → G as follows: Let a1 , a2 , . . . , an ∈ Q× satisfy
Nrd(a1 a2 · · · an ) = 1. Define θ by
(46.25) θ(e
a1e
a2 · · · e
an ) = ((a1 , a2 , . . . , an )).
The relation at the end of subsection 46.D and Lemmas 46.21, 46.22, and 46.23 show that
θ is a well defined homomorphism. Formulas (46.24) and (46.25) yield
Proposition 46.26. The maps π and θ are isomorphisms inverse to each other.
Corollary 46.27. The group A(Q)/A0 is generated by products ãb̃c̃ for all ordered
triples a, b, c of elements in Q× satisfying abc = 1 with the following set of defining
relations: ¡ ¢¡ ¢ ¡ ¢¡ ¢
e · (ab)c̃
(R10 ) ãb̃(cd) f · (da)
e d˜ = b̃c̃(da) f b̃c̃ for all a, b, c, d ∈ Q× such that abcd = 1;
(R20 ) ãb̃c̃ = 1 if a and b commute.
46.F. Proof of Proposition 46.6. To prove Proposition 46.6, we need to prove
that the homomorphism ∂1 in (46.10) is an isomorphism.
We shall view the fraction la /lb for a, b ∈ Q \ F as a nonzero rational function on C,
i.e., la /lb ∈ F (C)× .
Lemma 46.28. Let K0 be the quadratic subfield of Q corresponding to the fixed closed
point x0 of degree 2 on C and let b ∈ K0 \F . Then the space L1 consists of all the fractions
la /lb with a ∈ Q.
Proof. Obviously la /lb ∈ L1 . It follows from Lemma 45.5 that the space of all
fractions la /lb is 3-dimensional. On the other hand, dim L1 = 3. ¤
By Lemma 46.28, the group M 0 is generated by symbols of the form {la /lb , α} for all
a, b ∈ Q \ F and α ∈ F × and the group M1 is generated by symbols {la /lb , lc /ld } for all
a, b, c, d ∈ Q \ F .
Let a, b, c ∈ Q satisfy abc = 1. We define an element
[a, b, c] ∈ M1 /M 0
as follows: If at least one of a, b and c belongs to F × we set [a, b, c] = 0. Otherwise the
linear forms la , lb and lc are nonzero and we set
nl l o
a b
[a, b, c] := , + M 0.
lc lc
¡ ¢
Lemma 45.5 and the equality {u, −u} = 0 in K2 F (C) yield:
Lemma 46.29. Let a, b, c ∈ Q× be such that abc = 1 and let α ∈ F × . Then

(1) [a, b, c] = [b, c, a];
(2) [αa, α−1 b, c] = [a, b, c];
(3) [a, b, c] + [c−1 , b−1 , a−1 ] = 0;
(4) If a and b commute, then [a, b, c] = 0.
aebe
Lemma 46.30. ∂1 ([a, b, c]) = e c.
Proof. We may assume that none of a, b, or c is a constant. Let x, y, and z be the
points of C of degree 2 corresponding to quadratic subfields F [a], F [b], and F [c] that we
identify with F (x), F (y) and F (z) respectively.
Consider the following element in the class [a, b, c]:
nl l o nl o nl o
a b b b
w= , + , Nrd(a) + , − Nrd(b) .
lc lc lc la
By Proposition 45.12 (identifying residue fields with the corresponding quadratic ex-
tensions) and Lemma 45.5,
lb ¡ ¢−1 lb−1 ¡ ¢−1
∂x (w) = (x) − Nrd(b) = − Nrd(b) (x) − Nrd(b) = a,
lc lb−1 a−1
lc ¡ ¢ lb−1 a−1 ¡ ¢
∂y (w) = (y) − Nrd(ab) = − Nrd(a)−1 (x) − Nrd(ab)
la la−1
−1 −1
¡ ¢
= − Nrd(a) b̄ − Nrd(ab) = b,
la lbc
∂z (w) = − (z) Nrd(a)−1 = Nrd(a) (x) Nrd(a)−1 = c. ¤
lb lb
Lemma 46.31. Let a, b, c, d ∈ Q \ F be such that cd, da ∈
/ F and abcd = 1. Then
n l l lb ld o
a c
, ∈ M 0.
lcd lda lcd lda
Proof. In Proposition 45.8 plugging in the elements c−1 , ab and b for a, b and c
respectively and using Lemma 45.5, we get elements α, β, γ ∈ F × such that on the conic
C,
αla lc + βlb ld + γlcd lda = 0.
Then
αla lc βlb ld
− − =1
γlcd lda γlcd lda
and
n αla lc βlb ld o n la lc lb ld o
0= − ,− ≡ , mod M 0 . ¤
γlcd lda γlcd lda lcd lda lcd lda
Proposition 46.32. Let a, b, c, d ∈ Q× be such that abcd = 1. Then
[a, b, cd] + [ab, c, d] = [b, c, da] + [bc, d, a].
Proof. We first note that if one of the elements a, b, ab, c, d, cd belongs to F × , the
equality holds. For example, if a ∈ F × then the equality reads [ab, c, d] = [b, c, da] and
follows from Lemma 46.29 and if α = ab ∈ F × , then again by Lemma 46.29,
L.H.S. = 0 = [b, c, da] + [(da)−1 , α−1 c−1 , αb−1 ] = R.H.S.
So we may assume that none of the elements belong to F × . It follows from Lemma
45.5(4) that lcd /lab and lda /lbc belong to F × . By Lemmas 46.29 and 46.31, we have in
M1 /M 0 :
n l l lb ld o
a c
0= , + M0
lcd lda lcd lda
nl l o nl l o nl l o nl l o
a b c b a d c d
= , + , + , + , + M0
lcd lcd lda lda lcd lda lda lcd
³n l l o n l o´ n l l o
a d da ld c d
= [a, b, cd] − [b, c, da] + , + , + , + M0
lda lda lcd lda lda lcd
n l l o ³n l l o n l l o´
a d da d c d
= [a, b, cd] − [b, c, da] + , + , + , + M0
lbc lbc lcd lcd lda lcd
nl l o
c d
= [a, b, cd] − [b, c, da] − [bc, d, a] + , + M0
lcd lcd
nl l o
c d
= [a, b, cd] − [b, c, da] − [bc, d, a] + , + M0
lab lab
= [a, b, cd] − [b, c, da] − [bc, d, a] + [ab, c, d]. ¤
We shall use the presentation of the group A(Q)/A0 by generators and relations given
in Corollary 46.27. We define a homomorphism
µ : A(Q)/A0 → M1 /M 0
by the formula
aebe
µ(e c) := [a, b, c]
for all a, b, c ∈ Q such that abc = 1. It follows from Lemma 46.29(4) and Proposition
46.32 that µ is well defined. Lemma 46.30 implies that ∂1 ◦ µ is the identity.
To show that µ is the inverse of ∂1 it suffices to prove that µ is surjective.
The group M1 /M 0 is generated by elements of the form w = {la0 /lc0 , lb0 /lc0 } + M 0 for
a , b , c ∈ Q \ F . We may assume that 1, a0 , b0 and c0 are linearly independent (otherwise,
0 0 0
w = 0). In particular, 1, a0 , b0 and a0 b0 form a basis of Q, hence

c0 = α + βa0 + γb0 + δa0 b0
for some α, β, γ, δ ∈ F with δ 6= 0. We have
(γδ −1 + a0 )(β + δb0 ) = ε + c0
for ε = βγδ −1 − α. Set
a := γδ −1 + a0 , b := β + δb0 , c := (ε + c0 )−1 .
We have abc = 1. It follows from Lemma 45.5 that

nl 0 l 0 o nl l o
a b a b
w= , + M0 = , + M 0 = [a, b, c].
lc0 lc0 lc lc
By definition of µ, we have µ(e aebe
c) = [a, b, c] = w, hence µ is surjective. The proof of
Proposition 46.6 is complete. ¤
47. Hilbert theorem 90 for K2
In this section we prove the K2 -analog of the classical Hilbert Theorem 90.
Throughout this section let L/F be a Galois quadratic field extension with the Galois
group G = {1, σ}. For every field extension E/F linearly disjoint from L/F , the field
LE = L ⊗F E is a quadratic Galois extension of E with Galois group isomorphic to G.
The group G acts naturally on K2 (LE). We write (1 − σ)u for u − σ(u) with u ∈ K2 (LE)
and set
V (E) := K2 (LE)/(1 − σ)K2 (LE).
0
If E → E is a homomorphism of field extensions of F linearly disjoint from L/F ,
there is a natural homomorphism
V (E) → V (E 0 ).
Proposition 47.1. Let C be a conic curve over F and L/F a Galois quadratic field
extension such
¡ that
¢ C has a rational point over L. Then the natural homomorphism
V (F ) → V F (C) is injective.
¡ ¢
Proof. Let u ∈ K2 (L) satisfy uL(C) = (1 − σ)v for some v ∈ K2 L(C) . For a closed
point x ∈ C the L-algebra L(x) := L ⊗F F (x) is isomorphic to the direct product of
residue fields L(y) for all closed points y ∈ CL over x ∈ C. We denote the product of all
the ∂y (v) ∈ L(y)× with y over x by ∂x (v) ∈ L(x)× .
Set ax := ∂x (v) ∈ L(x)× . Then
¡ ¢ ¡ ¢ ¡ ¢
ax /σ(ax ) = ∂x (v)/σ ∂x (v) = ∂x (1 − σ)v = ∂x uL(C) = 1,
i.e., ax ∈ F (x)× . Applying Theorem 46.1 to CL , we have
Y ¡Y ¢ ¡Y ¢
NF (x)/F (ax ) = NL/F NL(y)/L (ay ) = NL/F NL(y)/L (∂y (v)) = 1.
x∈C y∈CL y∈CL
¢
¡
Hence applying Theorem 46.1 to C,¡ there¢ is a w ∈ K2 F (C) satisfying ∂x (w) = ax for
all x ∈ C. Set v 0 := v − wL(X) ∈ K2 L(C) . As
∂x (v 0 ) = ∂x (v)∂x (w)−1 = ax a−1
x = 1,
applying Theorem 46.1 to CL , there exists an s ∈ K2 (L) with sL(C) = v 0 . We have
(1 − σ)sL(C) = (1 − σ)v 0 = (1 − σ)v = uL(C) ,
i.e., (1 − σ)s − u splits over L(C). Since L(C)/L is a purely transcendental extension, we
must have (1 − σ)s − u = 0 (cf. Example 100.6) hence u = (1 − σ)s ∈ im(1 − σ). ¤
Corollary 47.2. For any finitely generated subgroup H ⊂ F × , there is a field exten-
sion F 0 /F linearly disjoint from L/F such that the natural homomorphism V (F ) → V (F 0 )
is injective and H ⊂ NL0 /F 0 (L0 × ) where L0 = LF 0 .
47. HILBERT THEOREM 90 FOR K2 213
Proof. By induction it suffices to assume that H is generated by one element b.

Set F 0 = F (C), where C = CQ is the conic curve associated with the quaternion algebra
Q = (L/F, b) (cf. §98.E). Since Q splits over F 0 , we have b ∈ NL0 /F 0 (L0 × ) by Facts 98.14(4)
and 98.15(5). The conic C has a rational point over L, therefore, the homomorphism
V (F ) → V (F 0 ) is injective by Proposition 47.1. ¤
For any two elements x, y ∈ L× , we write hx, yi for the class of the symbol {x, y} in
V (F ). Let f be the group homomorphism
¡ ¢
f = fF : NL/F (L× ) ⊗ F × → V (F ), f NL/F (x) ⊗ a = hx, ai.
The map f if well defined. Indeed, if NL/F (x) = NL/F (y) for x, y ∈ L× then y = xzσ(z)−1
for some z ∈ L× by the classical Hilbert theorem 90. Hence {y, a} = {x, a} + (1 − σ){z, a}
and consequently hy, ai = hx, ai.
¡ ¢
Lemma 47.3. Let b ∈ NL/F (L× ). Then f b ⊗ (1 − b) = 0.
Proof. If b = d2 for some d ∈ F × then
¡ ¢
f b ⊗ (1 − b) = hd, 1 − d2 i = hd, 1 − di + hd, 1 + di = h−1, 1 + di = 0
since −1 = zσ(z)−1 for some z ∈ L× .
So we may assume that b is not a square in F . Set
F 0 = F [t]/(t2 − b), L0 = L[t]/(t2 − b).
Note that L0 is either a field or product of two copies of the field F 0 . Let u ∈ F 0 be the class
of t, so that u2 = b. Choose x ∈ L× with NL/F (x) = b. Note that NL0 /F 0 (x/u) = b/u2 = 1
and NL0 /L (1 − u) = 1 − b.
The automorphism σ extends to an automorphism of L0 over F 0 . Applying the classical
Hilbert Theorem 90 to the extension L0 /F 0 , there is a v ∈ L0 × such that vσ(v)−1 = x/u.
We have
¡ ¢
f b ⊗ (1 − b) = hx, 1 − bi = hx, NL0 /L (1 − u)i = cF 0 /F hx, 1 − ui = cF 0 /F (hx/u, 1 − ui) =
cF 0 /F (hvσ(v)−1 , 1 − ui) = (1 − σ)cF 0 /F (hv, 1 − ui) = 0,

where cF 0 /F : V (F 0 ) → V (F ) is induced by the norm map cL0 /L : K2 (L0 ) → K2 (L). ¤
Theorem 47.4 (Hilbert Theorem 90 for K2 ). Let L/F be a Galois quadratic extension
and σ the generator of Gal(L/F ). Then the sequence
1−σ cL/F
K2 (L) −−→ K2 (L) −−→ K2 (F )
is exact.
Proof. Let u ∈ K2 (L) satisfy cL/F (u) = 0. By Proposition 100.2, the group K2 (L)
is generated by symbols of the form {x, a} with x ∈ L× and a ∈ F × . Therefore we can
write
Xm
u= {xj , aj }
j=1
for some xj ∈ L× and aj ∈ F × , and

m
X
cL/F (u) = {NL/F (xj ), aj } = 0.
j=1
Hence by definition of K2 (F ), we have in F × ⊗ F × :

m
X n
X ¡ ¢
(47.5) NL/F (xj ) ⊗ aj = ± bi ⊗ (1 − bi )
j=1 i=1
for some bi ∈ F × . Clearly, the equality (47.5) holds in H ⊗ F × for some finitely generated
subgroup H ⊂ F × containing all the NL/F (xj ) and bi .
By Corollary 47.2, there is a field extension F 0 /F linearly disjoint from L/F such
that the natural homomorphism V (F ) → V (F 0 ) is injective and H ⊂ NL0 /F 0 (L0 × ) with
L0 = LF 0 . The equality (47.5) then holds in NL0 /F 0 (L0 × ) ⊗ F 0 × . Now apply the map fF 0
to both sides of (47.5). By Lemma 47.3, the class of uL0 in V (F 0 ) is equal to
Xm ³X
m ´ X n
¡ ¢
hxj , aj i = fF 0 NL/F (xj ) ⊗ aj = ±fF 0 bi ⊗ (1 − bi ) = 0,
j=1 j=1 i=1
i.e., uL0 ∈ (1 − σ)K2 (L ). Since the map V (F ) → V (F 0 ) is injective, we conclude that

0
u ∈ (1 − σ)K2 (L). ¤
Theorem 47.6. Let u ∈ K2 (F ) satisfy 2u = 0. Then u = {−1, a} for some a ∈ F × .
In particular, u = 0 if char(F ) = 2.
Proof. Let G = {1, σ} be a cyclic group of order two. Consider a G-action on the
field L = F ((t)) of Laurent power series defined by
½
−t if char F 6= 2
σ(t) =
t/(1 + t) if char F = 2.
We have a Galois quadratic extension L/E with E = LG .
Consider the diagram
1−σ
K2 (L) −−−→ K2 (L)
 
 s
∂y y
{−1}
F × −−−→ K2 (F ),
where ∂ is the residue homomorphism of the canonical discrete valuation of L, the map
s = st is the specialization homomorphism of the parameter t (cf. §100.D), and the bottom
homomorphism is multiplication by {−1}. We claim that the diagram is commutative.
The group K2 (L) is generated by elements of the form {f, g} and {t, g} with f and g in
F [[t]] having nonzero constant term. If char F 6= 2, we have
s ◦ (1 − σ)({f, g}) = s({f, g} − {σ(f ), σ(g)})
= {f (0), g(0)} − {σ(f )(0), σ(g)(0)}
= 0 = {−1} · ∂({f, g})
48. PROOF OF THE MAIN THEOREM 215
and
s ◦ (1 − σ)({t, g}) = s({t, g} − {−t, σ(g)})
= {−1, g(0)}
= {−1} · ∂({t, g}).
¡ ¢
If char F = 2, we obviously have s(u) = s σ(u) for every u ∈ K2 (L), hence s◦(1−σ) =
0.
Since cL/E (uL ) = 2uE = 0, by Theorem 47.4, we have u = (1−σ)v for some v ∈ K2 (L).
The commutativity of the diagram yields
¡ ¢
u = s(uL ) = s (1 − σ)v = {−1, ∂(v)}. ¤
48. Proof of the main theorem

In this section we prove Theorem 44.1. Let F be a field of characteristic different from
two.
¡ ¢
48.A. Injectivity of hF . Suppose that hF u + 2K2 (F ) = 1 for an element u ∈
K2 (F ). Let u be a sum of n symbols. We prove by induction on n that µ u ∈ 2K ¶ 2 (F ).
a, b
First consider the case n = 1, i.e., u = {a, b} with a, b ∈ F × . Since is a split
F
quaternion algebra, there are x, y ∈ F such that b = x2 − ay 2 . It follows from Lemma
100.3 that {a, b} ∈ 2K2 (F ).
µ Next ¶ consider
µ ¶ the case n = 2, i.e. u =µ{a, b}¶+ {c, d}.
µ By ¶ assumption, the algebra
a, b c, d a, b c, d
⊗ is split, or equivalently, and are isomorphic. By the
F F F F
Chain Lemma 98.16, we may assume that a = c and hence u = {a, bd}, so the statement
follows from the case n = 1.
Now consider the general case. Write u in the form u = {a, b} + v for a, b ∈ F × and an
element v ∈ K2 F that is a sum ofµn − 1¶symbols. We may assume that {a, b} ∈ / 2K2 (F ),
a, b
i.e., the quaternion algebra Q := does not split. Let C = CQ be the conic curve
F
over F corresponding to Q and set L = F (C). The conic C is given by the equation
aX 2 + bY 2 = abZ 2
in projective coordinates (cf. Example 45.3). Set x = X/Z and y = Y /Z. As b−1 x2 +
a−1 y 2 = 1, we have
0 = {b−1 x2 , a−1 y 2 } = 2{x, a−1 y 2 } − 2{b, y} − {a, b}
and therefore {a, b} = 2r in K2 (L) with r = {x, a−1 y 2 } − {b, y}. Let p ∈ C be the point
of degree 2 given by Z = 0. The element r has only one nontrivial residue at the point p
and ∂p (r) = −1. µ ¶
a, b ¡ ¢
Since the quaternion algebra is split over L, we have hL vL + 2K2 L = 1. By
F
induction, vL = 2w for some element w ∈ K2 (L).
Set cx := ∂x (w) for every point x ∈ C. Since

c2x = ∂x (2w) = ∂x (vL ) = 1,
we have cx = (−1)nx for nx = 0 or 1. By Corollary 45.2(2), the degree of every point of
C is even, hence
X
nx deg(x) = 2m
x∈C
for some m ∈ Z. As every degree zero
Pdivisor on C is principal by Corollary 45.2(1), there
×
is a function f ∈ L with div(f ) = nx x − mp. Set
w0 := w + {−1, f } + kr ∈ K2 (L)
where k = m + np . If x ∈ C is a point different from p then
∂x (w0 ) = ∂x (w) · (−1)nx = 1.
Since also
∂p (w0 ) = ∂p (w) · (−1)m · (−1)k = (−1)np +m+k = 1,
we have ∂x (w0 ) = 1 for all x ∈ C. By Theorem 46.1, it follows that w0 = sL for some
s ∈ K2 (F ). Thus
vL = 2w = 2w0 − 2kr = 2sL − {ak , b}L .
Set v 0 := v − 2s + {ak , b} ∈ K2 (F ); we have 0
√ vL = 0. The conic C has a rational
point over the quadratic extension E = F ( a). Since the field extension E(C)/E is
0
purely transcendental and vE(C) = 0, we have vE0 = 0 by Example 100.6 and therefore
2v 0 = NE/F (vE0 ) = 0. It follows from Theorem 47.6 that v 0 = {−1, d} for some d ∈ F × .
Hence modulo 2K2 (F ), the element v is the sum of two symbols {ak , b} and {−1, d}.
Consequently, we are reduced to the case n = 2 that has already been considered.
48.B. Surjectivity of hF . Recall that we are writing k2 (F ) for K2 F/2K2 (F ).

Proposition 48.1. Let L/F be a quadratic extension. Then the sequence
rL/F cL/F
k2 (F ) −−−→ k2 (L) −−→ k2 (F )
is exact.
Proof. Let u ∈ K2 (L) satisfy cL/F (u) = 2v for some v ∈ K2 (F ). Then
cL/F (u − vL ) = 2v − 2v = 0
and, by Theorem 47.4, we have u − vL = (1 − σ)w for some w ∈ K2 (L). Hence
¡ ¢
u = vL + (1 − σ)w = v + cL/F (w) L − 2σw. ¤
We now finish the proof of Theorem 44.1. Let s ∈ Br2 (F ). First suppose that the field
F is 2-special (cf. §101.B). By induction on the index of s, we prove that s ∈ im(hF ).
By Proposition 101.15, there exists a quadratic extension L/F with ind(sL ) < ind(s). By
induction, sL = hL (u) for some u ∈ k2 (L). It follows from Proposition 101.9 that
¡ ¢ ¡ ¢
hF cL/F (u) = cL/F hL (u) = cL/F (sL ) = 1.
48. PROOF OF THE MAIN THEOREM 217
The injectivity of hF implies that cL/F (u) = 0 so by Proposition 48.1, we have u = vL for
some v ∈ k2 (F ). Therefore
hF (v)L = hL (vL ) = hL (u) = sL
hence s−hF (v) splits over L and must be the class of a quaternion algebra. Consequently,
s − hF (v) = hF (w) for some symbol w in k2 (F ), so s = hF (v + w) ∈ im(hF ).
In the general case, applying the first part of the proof to a maximal odd degree
extension of F (cf. §101.B and Proposition 101.16), we see that there exists an odd degree
extension E/F such that sE = hE (v) for some v ∈ k2 (E). Then again by Proposition
101.9, ¡ ¢ ¡ ¢
s = cE/F (sE ) = cE/F hE (v) = hF cE/F (v) . ¤
Part
Algebraic cycles
CHAPTER IX
Homology and cohomology
The word “scheme” in this book (with the exception of §49) will mean a separated
scheme of finite type over a field and a “variety” will mean an integral scheme.
In this chapter we develop the K-homology and K-cohomology theories of schemes
over a field that generalizes the theory of Chow groups. We follow the approach of [119]
given by Rost. There are two advantages of having such general theories rather than just
that of Chow groups. First we have a long (infinite) localization exact sequence. This
tool together with the 5-lemma allows us to give simple proofs of some basic results in the
theory such as the Homotopy Invariance and Projective Bundle Theorems. Secondly, the
construction of the deformation map (called the specialization homomorphism in [47]),
used in the definition of the pull-back homomorphisms, is much easier – it does not require
intersections with Cartier divisors.
We view the K-homology as a covariant functor from the category of schemes of
finite type over a field to the category of abelian groups and the K-cohomology as a
contravariant functor from the category of smooth schemes of finite type over a field. The
fact that K-homology groups for smooth schemes coincide with K-cohomology groups
should be viewed as Poincaré duality.
49. The complex C∗ (X)

The purpose of this section is to construct complexes C∗ (X) giving the homology and
cohomology theories that we need.
Throughout this section (and only in this section), we need to extend the class of
separated scheme of finite type over a field and consider the class of excellent schemes
of finite dimension. A noetherian scheme X is called excellent if the local ring OX,x
is excellent for every x ∈ X [100]. The class of excellent schemes of finite dimension
contains:
1. Schemes of finite type over a field.
2. Closed
¡ and¢ open subschemes of excellent schemes.
3. Spec OX,x where x is a point of a scheme X of finite type over a field.
4. Spec(R) where R is a complete noetherian local ring.
We shall use the following properties of excellent schemes:
A. If X is excellent integral then the normalization morphism Xe → X is finite and X
e
is excellent.
B. An excellent scheme X is catenary, i.e., given irreducible closed subschemes Z ⊂
Y ⊂ X, all maximal chains of closed irreducible subsets between Z and Y have the same
length.
221
222 IX. HOMOLOGY AND COHOMOLOGY
b is its completion then the induced morphism

C. If R is a local excellent ring and R
b → Spec(R) is flat.
Spec(R)
If x is a point of a scheme X, we write κ(x) for the residue field of x (and we shall use
the standard notation F (x) when X is a scheme over a field F ). We write dim x for the
dimension of the closure {x} and X(p) for the set of point of X of dimension p.
An integral scheme is called a variety.
49.A. Residue homomorphism for local rings. Let R be a 1-dimensional local
excellent domain with quotient field L and residue field E. Let R e denote the integral
closure of R in L. The ring R e is semilocal, 1-dimensional, and finite as R-algebra. Let
M1 , M2 , . . . , Mn be all of the maximal ideals of R.e Each localization R eM is integrally
i
closed, noetherian and 1-dimensional hence a DVR. Let vi denote the discrete valuation
eM and Ei its residue field. The field extension Ei /E is finite. Let K∗ denote the
of R i
Milnor K-groups (cf. §100) and define the residue homomorphism
∂R : K∗ (L) → K∗−1 (E),
by the formula
n
X
∂R = cEi /E ◦ ∂vi ,
i=1
where
∂vi : K∗ (L) → K∗−1 (Ei )
is the residue homomorphism associated with the discrete valuation vi on L (cf. §100.B)
and
cEi /E : K∗−1 (Ei ) → K∗−1 (E)
is the norm homomorphism (cf. §100.E).
Let X be an excellent scheme. For every pair of points x, x0 ∈ X, we define a homo-
morphism ¡ ¢ ¡ ¢
∂xx0 : K∗ κ(x) → K∗−1 κ(x0 )
as follows: Let Z be the closure of {x} in X considered as an integral closed subscheme
(subvariety) of X. If x0 ∈ Z (in this case we say that x0 is a specialization of x) and
dim x = dim x0 + 1, then the local ring R = OZ,x0 is a 1-dimensional excellent local
domain with quotient field κ(x) and residue field κ(x0 ). We set ∂xx0 = ∂R . Otherwise set
∂xx0 = 0.
Lemma 49.1.
¡ Let
¢ X be an excellent scheme of finite dimension. For each x ∈ X and
every α ∈ K∗ κ(x) the residue ∂x0 (α) is nontrivial for only finitely many points x0 ∈ X.
x
Proof. We may assume that X = Spec(A) where A is an integrally closed domain,

x the generic point of X and α the symbol {a1 , a2 , . . . , an } with nonzero ai ∈ A. For
every point x0 ∈ X of codimension 1, let vx0 be the corresponding discrete valuation of
the quotient field of A. For each i, there is a bijection between the set of all x0 satisfying
vx0 (ai ) 6= 0 and the set of minimal prime ideals of the (noetherian) ring A/ai A and hence
this set is finite. Thus, for all but finitely many x0 we have vx0 (ai ) = 0 for all i and
therefore ∂xx0 (α) = 0. ¤
49. THE COMPLEX C∗ (X) 223
It follows from Lemma 49.1 that there is a well defined endomorphism d = dX of the
direct sum a ¡ ¢
C(X) := K∗ κ(x)
x∈X
such that the (x, x )-component of d is equal to ∂xx0 .
0
Example 49.2. Let Z ⊂ X be a closed subscheme and let z1 , z2 , . . . be all of the

generic points of Z. We set
X a ¡ ¢
[Z] := mi zi ∈ K0 κ(x) ⊂ C(X),
x∈X
where mi = l(OZ,zi ) is the length of the local ring OZ,zi . The element [Z] is called the
cycle of Z on X.
Example 49.3. Let X be an excellent¡ scheme
¢ of finite dimension, x ∈ X, and f ∈
κ(x)× . We view f as an element of K1 κ(x) ⊂ C(X). Then the element
a ¡ ¢
dX (f ) ∈ K0 κ(x) ⊂ C(X)
x∈X
is called the divisor of f and is denoted by div(f ) .

The group C(X) is graded: we write for any p ≥ 0,
a ¡ ¢
Cp (X) := K∗ κ(x) .
x∈X(p)
The endomorphism dX of C∗ (X) has degree −1 with respect to this grading. We also set
a ¡ ¢
Cp,n (X) := Kp+n κ(x) ,
x∈X(p)
hence Cp (X) is the coproduct of Cp,n (X) over all n. Note that the graded group C∗,n (X)
is invariant under dX for every n.
Let X be a scheme over a field F . Then the group Cp (X) has a natural structure of a
left and right K∗ (F )-module for all p and dX is a homomorphism of right K∗ (F )-modules.
If X is the disjoint union of two schemes X1 and X2 , we have
(49.4) C∗ (X) = C∗ (X1 ) ⊕ C∗ (X2 )
and dX = dX1 ⊕ dX2 .
49.B. Multiplication with an invertible function. Let a be an invertible regular
function on an excellent scheme X. For every α ∈ C∗ (X), we write {a} · α for the element
of C∗ (X) satisfying
({a} · α)x = {a(x)} · αx
for every x ∈ X. We denote by {a} the endomorphism of C∗ (X) given by α 7→ {a} · α.
The product α · {a} is defined similarly.
Let a1 , a2 , . . . , an be invertible regular functions on an excellent scheme X. We write
{a1 , a2 , . . . , an } · α for the product {a1 } · {a2 } · . . . · {an } · α and {a1 , a2 , . . . , an } for the
endomorphism of C∗ (X) given by α 7→ {a1 , a2 , . . . , an } · α.
Proposition 49.5. Let a be an invertible function on an excellent scheme X and

α ∈ C∗ (X). Then
¡ ¢ ¡ ¢
dX α · {a} = dX (α) · {a} and dX {a} · α = −{a} · dX (α).
Proof. The statement follows from Fact 100.4(1) and the projection formula for the
norm map in Proposition 100.8(3). ¤
By Fact 100.1, it follows that
{a1 , a2 } = −{a2 , a1 } and {a1 , a2 } = 0 if a1 + a2 = 1.
49.C. Push-forward homomorphisms. Let f : X → Y be a morphism of excel-
lent schemes. We define the push-forward homomorphism
f∗ : C∗ (X) → C∗ (Y )
as follows: Let x ∈ X and y ∈ Y . If y = f (x) ∈ Y and the field extension κ(x)/κ(y) is
finite, we set ¡ ¢ ¡ ¢
(f∗ )xy := cκ(x)/κ(y) : K∗ κ(x) → K∗ κ(y)
and set (f∗ )xy = 0 otherwise. It follows from transitivity of the norm map that if g : Y → Z
is another morphism then (g ◦ f )∗ = g∗ ◦ f∗ .
If either
(1) f is a morphism of schemes of finite type over a field or
(2) f is a finite morphism,
the push-forward f∗ is a graded homomorphism of degree 0. Indeed if y = f (x) then
dim y = dim x if and only if κ(x)/κ(y) is a finite extension for all x ∈ X.
If f is a morphism of schemes over a field F then f∗ is a homomorphism of left and
right K∗ (F )-modules.
Example 49.6. If f : X → Y is a closed embedding then f∗ is a monomorphism
satisfying f∗ ◦ dX = dY ◦ f∗ . Moreover, if in addition f is a bijection on points (e.g., if f
is the canonical morphism Yred → Y ) then f∗ is an isomorphism.
Remark 49.7. Let X be a localization of a scheme Y (e.g., X is an open subscheme
of Y ) and f : X → Y the natural morphism. For every point x ∈ X, the natural ring
homomorphism OY,f (x) → OX,x is an isomorphism. It follows from the definitions that
for any x, x0 ∈ X, we have
0
(f∗ ◦ dX )xy0 = (f∗ )xy0 ◦ (dX )xx0 = (dY )yy0 ◦ (f∗ )xy = (dY ◦ f∗ )xy0
where y = f (x) and y 0 = f (x0 ). Note that if y 00 ∈ Y does not belong to the image of f
then (f∗ ◦ dX )xy00 = 0 but in general (dY ◦ f∗ )xy00 may be nonzero.
The following rule is a consequence of the projection formula for Milnor’s K-groups
(cf. Fact 100.8(3)).
Proposition 49.8. Let f : X → Y be a morphism of schemes and a an invertible
regular function on Y . Then
f∗ ◦ {a0 } = {a} ◦ f∗
where a0 = f ∗ (a) = a ◦ f .
Proposition 49.9. Let f : X → Y be either
(1) a proper morphism of schemes of finite type over a field or

(2) a finite morphism.
Then the diagram
X d
Cp (X) −−−→ Cp−1 (X)
 

f∗ y
f
y∗
Y d
Cp (Y ) −−−→ Cp−1 (Y )
is commutative.
Proof. Let x ∈ X(p) and y 0 ∈ Y(p−1) . The (x, y 0 )-component of both compositions in
the diagram can be nontrivial only if y 0 belongs to the closure of the point y = f (x), i.e.,
if y 0 is a specialization of y. We have
p = dim x ≥ dim y ≥ dim y 0 = p − 1,
therefore, dim y can be either equal to p or p − 1. Note that if f is finite then dim y = p.
Case 1: dim(y) = p.
In this case, the field extension κ(x)/κ(y) is finite. Replacing X by the closure of {x}
and Y by the closure of {y} we may assume that x and y are the generic points of X and
Y respectively.
First suppose that X and Y are normal. Since the morphism f is proper, the points
x0 ∈ Xp−1 satisfying f (x0 ) = y 0 are in a bijective correspondence with the extensions of
the valuation vy0 of the field κ(y) to the field κ(x). Hence by Fact 100.8(4),
(dY ◦ f∗ )xy0 =∂yy0 ◦ cκ(x)/κ(y)

X
= cκ(x0 )/κ(y0 ) ◦ ∂xx0
f (x0 )=y 0
X 0
= (f∗ )xy0 ◦ (dX )xx0
f (x0 )=y 0
=(f∗ ◦ dX )xy0 .
e → X and h : Ye → Y be the normalizations and let x̃ and

In the general case let g : X
ỹ be the generic points of Xe and Ye respectively. Note that κ(x̃) ' κ(x) and κ(ỹ) ' κ(y).
There is a natural morphism f˜ : X e → Ye over f .
Consider the following diagram
¡ ¢ dXe
K∗ κ(x̃) WW /C e
p−1 (X) WWWW
WWWWW WWWWW
WWWWW ∼ WWWWWg∗
WWW WWWWW W WWWWW
WWWWW WWWWW
WW+ ¡ ¢ dX
WWW+
cκ(x̃)/κ(ỹ) fe∗ K∗ κ(x) / Cp−1 (X)
¡² ¢ dYe
²
K∗ κ(ỹ) WW /C e
p−1 (Y ) WWWW
cκ(x)/κ(y) f∗
WWWWW WW W
WWWWW ∼ WWWWW h∗
WWWWW WWWWW
WWWWW WWWWW
WWWWW WWWWW
+ ¡² ¢ dY W+ ²
K∗ κ(y) / Cp−1 (Y )
By the first part of the proof, the back face of the diagram is commutative. The left face
is obviously commutative. The right face is commutative by functoriality of the push-
forward. The upper and the bottom faces are commutative by definition of the maps
dX and dY . Hence the front face is also commutative, i.e., the (x, y 0 )-components of the
compositions f∗ ◦ dX and dY ◦ f∗ coincide.
Note that we have proved the proposition in the case when f is finite. Before proceed-
ing to Case 2, as a corollary we also deduce
Theorem 49.10 (Weil’s Reciprocity Law). Let X be a complete integral curve over a
field F . Then the composition
¡ ¢ dX a ¡ ¢ P cκ(x)/F
K∗+1 F (X) −→ K∗ F (x) −−−−−→ K∗ (F )
x∈X(0)
is trivial.
Proof. The case X = P1F follows from Theorem 100.7. The general case can be
reduced to the case of the projective line as follows. Let f be a nonconstant rational
function on X. We view f as a finite morphism f : X → P1F over F . By the first case of
the proof of Proposition 49.9, the left square of the diagram
¡ ¢ dX ` ¡ ¢ P cκ(x)/F
K∗+1 F (X) −−−→ x∈X(0) K∗ F (x) −−−−−−→ K∗ (F )
  °
f f °
y∗ y∗ °
P
¡ ¢ dP1 ` ¡ ¢ cκ(y)/F
K∗+1 F (P1 ) −−− → y∈P1 K ∗ F (y) −−−−−→ K∗ (F )
(0)
is commutative. The right square is commutative by the transitivity property of the

norm map (cf. Fact 100.8(1)). Finally, the statement of the theorem follows from the
commutativity of the diagram. ¤
Weil’s Reciprocity Law can be reformulated as follows:
Corollary 49.11. Proposition 49.9 holds for the structure morphism X → Spec(F ).
We return to the proof of Proposition 49.9.

Case 2: dim(y) = p − 1. ¡ ¢ ¡ ¢
In this case y 0 = y. We replace Y by Spec κ(y) and X by the fiber X ×Y Spec κ(y) of
f over y. We can further replace X by the closure of x in X. Thus, X is a proper integral
curve over the field κ(y) and the result follows from Corollary 49.11. ¤
49.D. Pull-back homomorphisms. Let g : Y → X be a flat morphism of excellent
schemes. We say that g is of relative dimension d if for every x ∈ X in the image of g
and for every generic point y of g −1 ({x}) we have dim y = dim x + d.
In this book all flat morphisms will be assumed of constant relative dimension.
Let g : Y → X be a flat morphism of relative dimension d. For every point x ∈ X, let
Yx denote the fiber scheme ¡ ¢
Y ×X Spec κ(x)
over κ(x). We identify the underlying topological space of Yx with a subspace of Y .
The following statement is a consequence of the going-down theorem [100, Ch. 1, Th.
4].
Lemma 49.12. For every x ∈ X we have:
(1) dim y ≤ dim x + d for every y ∈ Yx .
(2) A point y ∈ Yx is generic in Yx if and only if dim y = dim x + d.
If y is a generic point of Yx , the local ring OYx ,y is noetherian 0-dimensional hence is
artinian. We define the ramification index of y by
ey (g) := l(OYx ,y ),
where l denotes the length (cf. §102).
The pull-back homomorphism
g ∗ : C∗ (X) → C∗+d (Y )
is defined as follows: Let x ∈ X and y ∈ Y . If g(y) = x and y is a generic point of Yx , we
set ¡ ¢ ¡ ¢
(g ∗ )xy := ey (g) · rκ(y)/κ(x) : K∗ κ(x) → K∗ κ(y)
where rκ(y)/κ(x) is the restriction homomorphism (cf. §100.A) and (g∗ )xy = 0 otherwise.
Suppose that Z is of pure dimension d over a field F . The structure morphism p : Z →
Spec(F ) is flat of relative dimension d. The image of the identity under the composition
p∗
p∗ : Z = K0 (F ) = C0,0 (Spec F ) − ∗i
→ Cd,−d (Z) −
→ Cd,−d (X),
where i : Z → X is the closed embedding, is equal to the cycle [Z] of Z.
Example 49.13. Let p : E → X be ¡ a ¢vector bundle of rank r. Then p is a flat
morphism of relative dimension r and p∗ [X] = [E].
Example 49.14. Let X be a scheme of finite type over F and let L/F be an arbitrary
field extension. The natural morphism g : XL → X is flat of relative dimension 0. The
pull-back homomorphism
g ∗ : Cp (X) → Cp (XL )
is called the change of field homomorphism.
Example 49.15. An open embedding j : U → X is a flat morphism of relative

dimension 0. The pull-back homomorphism
j ∗ : Cp (X) → Cp (U )
is called the restriction homomorphism.
The following proposition is an immediate consequence of the definitions.
Proposition 49.16. Let g : Y → X be a flat morphism and a an invertible function
on X. Then
g ∗ ◦ {a} = {a0 } ◦ g ∗ ,
where a0 = g ∗ (a) = a ◦ g.
Let g be a morphism of schemes over a field F . It follows from Proposition 49.16 that
g ∗ is a homomorphism of left and right K∗ (F )-modules.
Let g : Y → X and h : Z → Y be flat morphisms. Let z ∈ Z and y = h(z), x = g(y).
It follows from Lemma 49.12 that z is a generic point of Zx if and only if z is a generic
point of Zy and y is a generic point of Yx .
Lemma 49.17. Let z be a generic point of Zx . Then ez (g ◦ h) = ez (h) · ey (g).
Proof. The statement follows from Corollary 102.2 with B = OYx ,y and C = OZx ,z .
Note that C/QC = OZy ,z where Q is the maximal ideal of B. ¤
Proposition 49.18. Let g : Y → X and h : Z → Y be flat morphisms of constant
relative dimension. Then (g ◦ h)∗ = h∗ ◦ g ∗ .
Proof. Let x ∈ X and z ∈ Z. We compute the (z, x)-components of both sides of
the equality. We may assume that x = (g ◦ h)(z). Let y = h(z). By Lemma 49.17, we
have
¡ ¢x
(g ◦ h)∗ z = ez (g ◦ h) · rκ(z)/κ(x)
= ez (h) · ey (g) · rκ(z)/κ(y) ◦ rκ(y)/κ(x)
= (h∗ )yz ◦ (g ∗ )xy
= (h∗ ◦ g ∗ )xz . ¤
Next consider the fiber product diagram
g0
X 0 −−−→ X
 
(49.19) 
f 0y
f
y
g
Y 0 −−−→ Y.
Proposition 49.20. Let g and g 0 in (49.19) be flat morphisms of relative dimension

d. Suppose that either
(1) f is a morphism of schemes of finite type over a field or
(2) f is a finite morphism.
Then the diagram

g0 ∗
Cp (X) −−−→ Cp+d (X 0 )
 

f∗ y
f 0
y∗
g∗
Cp (Y ) −−−→ Cp+d (Y 0 )
is commutative.
Proof. Let x ∈ X(p) and y 0 ∈ Y(p+d)
0
. We shall compare the (x, y 0 )-components of
both compositions in the diagram. These components are trivial unless g(y 0 ) = f (x).
Denote this point by y. By Lemma 49.12,
p + d = dim y 0 ≤ dim y + d ≤ dim x + d = p + d,
hence dim y = dim x = p and y 0 is a generic point of Yy0 . In particular, the field extension
κ(x)/κ(y) is finite.
Let S be the set of all x0 ∈ X 0 satisfying f 0 (x0 ) = y 0 and g 0 (x0 ) = x. Again by Lemma
49.12,
p + d = dim y 0 ≤ dim x0 ≤ dim x + d = p + d,
hence dim x0 = dim y 0 = p + d and x0 is a generic point of Xx0 . In particular, the field
extension κ(x0¡)/κ(y 0 ) is finite. ¢The set S is in a natural bijective correspondence with the
finite set Spec κ(y 0 ) ⊗κ(y) κ(x) .
The local ring C = OXx0 ,x0 is a localization of the ring OYy0 ,y0 ⊗κ(y) κ(x) and hence is
flat over B = OYy0 ,y0 . Let Q be the maximal ideal of B. The factor ring C/QC is the
localization of the tensor product κ(y 0 ) ⊗κ(y) κ(x) at the prime ideal corresponding to x0 .
Denote by lx0 the length of C/QC.
By Corollary 102.2,
(49.21) ex0 (g 0 ) = lx0 · ey0 (g)
for every x0 ∈ S. It follows from (49.21) and Fact 100.8(5) that
∗
X 0 ∗
(f∗0 ◦ g 0 )xy0 = (f∗0 )xy0 ◦ (g 0 )xx0
x0 ∈S
X
= ex0 (g 0 ) · cκ(x0 )/κ(y0 ) ◦ rκ(x0 )/κ(x)
x0 ∈S
X
= ey0 (g) · lx0 · cκ(x0 )/κ(y0 ) ◦ rκ(x0 )/κ(x)
x0 ∈S
= ey0 (g) · rκ(y0 )/κ(y) ◦ cκ(x)/κ(y)
= (g ∗ )yy0 ◦ (f∗ )xy
= (g ∗ ◦ f∗ )xy0 ¤
Remark 49.22. It follows from the definitions that Proposition 49.20 holds also for
arbitrary f if Y 0 is a localization of Y (cf. Remark 49.7).
Proposition 49.23. Let g : Y → X be a flat morphism of relative dimension d. Then

the diagram
X d
Cp (X) −−−→ Cp−1 (X)
 

g∗ y
g∗
y
Y d
Cp+d (Y ) −−−→ Cp+d−1 (Y )
is commutative.
Proof. Let x ∈ X(p) and y 0 ∈ Y(p+d−1) . We compare the (x, y 0 )-components of both
compositions in the diagram. Let y1 , . . . , yk be all generic points of Yx ⊂ Y satisfying
y 0 ∈ {yi }. We have
k
X k
X
yi
(49.24) (dY ◦ g ∗ )xy0 = ∗ x
(dY )y0 ◦ (g )yi = eyi (g) · (dY )yyi0 ◦ rκ(yi )/κ(x) .
i=1 i=1
Set x0 = g(y 0 ). If x0 ∈
/ {x}, then both components (g ∗ ◦ dX )xy0 and (dY ◦ g ∗ )xy0 are trivial.
Suppose x0 ∈ {x}. We have
p = dim x ≥ dim x0 ≥ dim y 0 − d = p − 1.
Therefore, dim x0 is either p or p − 1.
Case 1: dim(x0 ) = p, i.e., x0 = x.
The component (g ∗ ◦ dX )xy0 is trivial since (g ∗ )x̃y0 = 0 for every x̃ 6= x0 . By assumption,
every discrete valuation of κ(yi ) with center y 0 is trivial on κ(x). Therefore the map (dY )yyi0
is trivial on the image of rκ(yi )/κ(x) . It follows from formula (49.24) that (dY ◦ g ∗ )xy0 = 0.
Case 2: dim(x0 ) = p − 1.
We have y 0 is a generic point of Yx0 and
0
(49.25) (g ∗ ◦ dX )xy0 = (g ∗ )xy0 ◦ (dX )xx0 = ey0 (g) · rκ(y0 )/κ(x0 ) ◦ ∂xx0 .
Replacing X by {x} and Y by g −1 ({x}), we may assume that X = {x}. By Proposi-
tions 49.9 and 49.20, we can replace X by its normalization X e and Y by the fiber product
Y ×X X,e so we may assume that X is normal.
Let Y1 , . . . , Yk be all irreducible components of Y containing y 0 , so that yi is the generic
point of Yi for all i. Let Yei be the normalization of Yi and let ỹi be the generic points of Yei .
We have κ(ỹi ) = κ(yi ). Let t be a prime element of the discrete valuation ring R = OX,x0 .
The local ring A = OY,y0 is one-dimensional; its minimal prime ideals are in a bijective
correspondence with the set of points y1 , . . . , yk .
Fix i ∈ [1, k]. We write Ai for the factor ring of A by the corresponding minimal prime
ideal. Since A is flat over R, the prime element t is not a zero divisor in A, hence the
image of t in Ai is not zero for every i. Let A ei be the normalization of the ring Ai .
e
Let Si be the set of all points w ∈ Yi such that g(w) = x0 . There is a natural bijection
between Si and the set of all maximal ideals of A ei . Moreover, if Q is a maximal ideal of A ei
corresponding to a point w ∈ Si then the local ring OYei ,w coincides with the localization
ei with respect to Q.
of A
Denote by li,w the length of the ring OYei ,w /tOYei ,w . Applying Lemma 102.3 to the
ei and M = A
A-algebra A ei /tAei , we have
¡ ¢ X £ ¤
(49.26) lA A ei /tA
ei = li,w · κ(w) : κ(y 0 ) .
w∈Si
On the other hand, li,w is the ramification index of the discrete valuation ring OYei ,w
over R. It follows from Fact 100.4(2) that
(49.27) ∂wỹi ◦ rκ(yi )/κ(x) = li,w · rκ(w)/κ(x0 ) ◦ ∂xx0
for every w ∈ Si .
By (49.26), (49.27), and Fact 100.8(3), for every i, we have
X
(dY )yyi0 ◦ rκ(yi )/κ(x) = cκ(w)/κ(y0 ) ◦ ∂wỹi ◦ rκ(yi )/κ(x)
X
= cκ(w)/κ(y0 ) · li,w · rκ(w)/κ(x0 ) ◦ ∂xx0
X
= li,w · cκ(w)/κ(y0 ) ◦ rκ(w)/κ(y0 ) ◦ rκ(y0 )/κ(x0 ) ◦ ∂xx0
X £ ¤
= li,w · κ(w) : κ(y 0 ) · rκ(y0 )/κ(x0 ) ◦ ∂xx0
¡ ¢
= lA Aei /tA
ei · rκ(y0 )/κ(x0 ) ◦ ∂ x0
x
(where all summations are taken over all w ∈ Si ).

The factor A-module A ei /Ai is of finite length hence by Lemma 102.4, we have h(t, Ai ) =
ei ) where h is the Herbrand index. Since t is not a zero divisor in either Ai or in A
h(t, A ei ,
we have lA (Aei /tA
ei ) = lA (Ai /tAi ) = l(Ai /tAi ). Therefore
(49.28) (dY )yyi0 ◦ rκ(yi )/κ(x) = l(Ai /tAi ) · rκ(y0 )/κ(x0 ) ◦ ∂xx0 .
The local ring OYx ,yi = OY,yi is the localization of A with respect to the minimal prime
ideal corresponding to yi . The ring OYx0 ,y0 is canonically isomorphic to A/tA.
Applying Lemma 102.5 to the ring A and the module M = A we get the equality
k
X k
X
(49.29) ey0 (g) = h(t, A) = l(OYx ,yi ) · l(Ai /tAi ) = eyi (g) · l(Ai /tAi ).
i=1 i=1
It follows from (49.25), (49.28), and (49.29) that

k
X
(dY ◦ g ∗ )xy0 = eyi (g) · (dY )yyi0 ◦ rκ(yi )/κ(x)
i=1
k
X
= eyi (g) · l(Ai /tAi ) · rκ(y0 )/κ(x0 ) ◦ ∂xx0
i=1
= ey0 (g) · rκ(y0 )/κ(x0 ) ◦ ∂xx0
= (g ∗ ◦ dX )xy0 . ¤
The following proposition was proven by Kato in [78].
Proposition 49.30. For every excellent scheme X over a field, the map dX is a
differential of C∗ (X), i.e., (dX )2 = 0.
Proof. We will prove the statement in several steps.

Step 1: X = Spec(R), where R = F [[s, t]] and F is a field.
A polynomial tn + a1 tn−1 + a2 tn−2 + · · · + an over the ring F [[s]] is called marked if
ai ∈ sF [[s]] for all i. We shall use the following properties of marked polynomials derived
from the Weierstrass Preparation Theorem [20, Ch.VII, §3, no 8]:
A. Every height 1 ideal of the ring R is either equal to sR or is generated by a unique
marked polynomial.
B. A marked polynomial f is irreducible in R if and only if f is irreducible in F ((s))[t].
It follows that the multiplicative group F ((s, t))× is generated by R× , s, t and the set
H of all power series of the
¡ form t−n
¢ · f where f is a marked polynomial of degree n.
×
If r ∈ R and α ∈ K∗ F ((s, t)) then by Proposition 49.5,
¡ ¢ ¡ ¢
(dX )2 {r} · α = −dX {r̄} · dX (α) = {r̄} · (dX )2 (α),
where r̄ ∈ F ¡is the residue

¢ of r. Thus it suffices to prove the following:
(i) (dX )2 {s,
¡ t} = 0, ¢
(ii) (dX )2 {f, g1 , . . . , gn } = 0 where f ∈ H and all gi belong to the subgroup
generated by s, t and H.
For every point x ∈ X(1) set ∂x = ∂xy , where y is the generic point of X and ∂ x = ∂zx ,
where z is the closed point of X. Thus,
¡ ¢y X ¡ ¢
(dX )2 z = ∂ x ◦ ∂x : K∗ F ((s, t)) → K∗−2 (F ).
x∈X(1)
To prove (i) let xs and xt be the points of X(1) given by the ideals sR and tR respec-
tively. We have
X ¡ ¢ ¡ ¢ ¡ ¢
∂ x ◦ ∂x {s, t} = ∂ xs {t} − ∂ xt {s} = 1 − 1 = 0.
x∈X(1)
To prove (ii) consider the field L = F ((s)) and the natural morphism
¡ ¢ ¡ ¢
h : X 0 = Spec R[s−1 ] → Spec L[t] = A1L .
0
By the properties of marked polynomials, the map h identifies the set X(0) = X(1) − {xs }
with the subset of the closed points of AL given by irreducible marked polynomials. For
1
every x ∈ X 0 we write x̄ for the point h(x) ∈ A1L . Note that for x ∈ X(0)
0
= X(1) − {xs },
the residue fields κ(x) and L(x̄) are canonically isomorphic. In particular, the field κ(x)
can be viewed as a finite extension of L. By Fact 100.8(4), we have ∂ x = ∂ ◦ cκ(x)/L , where
∂ : K∗ (L) → K∗−1 (F ) is given by the canonical discrete valuation of L.
Let x ∈ X(0)0
= X(1) − {xs }. We write ∂x̄ for ∂x̄ȳ . Under the identification of κ(x)
¡ ¢ ¡ ¢
with L(x̄), we have ∂x̄ = ∂x ◦ i where i : K∗ L(t) → K∗ F ((s, t)) is the canonical
homomorphism. Therefore,
X X
∂ x ◦ ∂x ◦ i = ∂ xs ◦ ∂xs ◦ i + ∂ ◦ cκ(x)/L ◦ ∂x ◦ i
x∈X(1) 0
x∈X(0)
X
= ∂ xs ◦ ∂xs ◦ i + ∂ ◦ cL(x̄)/L ◦ ∂x̄ .
0
x∈X(0)
¡ ¢
Let α = {f, g1 , . . . , gn } ∈ Kn+1 L(t) with f and gi as in (ii). Note that the divisors
in A1L of the functions f and gi are supported in the image of h. Hence ∂p (α) = 0 for
every closed point A1L not in the image of h. Moreover, for the point q of P1L at infinity,
f (q) = 1 and therefore, ∂q (α) = 0. Hence, by Weil’s Reciprocity Law 49.10, applied to
P1L , X X
cL(x̄)/L ◦ ∂x̄ (α) = cL(p)/L ◦ ∂p (α) = 0.
0
x∈X(0) p∈P1L
Notice also that f (xs ) = 1 hence ∂xs ◦ i(α) = 0 and therefore,

¡ ¢ X
(dX )2 {f, g1 , . . . , gn } = ∂ x ◦ ∂x ◦ i(α) = 0.
x∈X(1)
Step 2: X = Spec(S), where S is a (noetherian) local complete two-dimensional ring

containing a field.
Let M ⊂ S be the maximal ideal. By Cohen’s theorem [147, Ch. VIII, Th.27], there is a
subfield F ⊂ S such that the natural ring homomorphism F → S/M is an isomorphism.
Choose local parameters s, t ∈ M and consider the subring R = F [[s, t]] ⊂ S. Denote
by P the maximal ideal of R. There is an integer r such that M r ⊂ P S. We claim that
the R-algebra S is finite. Indeed, note that
\ \
P nS ⊂ M n = 0.
n>0 n>0
Since S/M is of finite length and there is a natural surjection S/M r → S/P S, the ring
r
S/P S is a finitely generated R/P -module. As the ring R is complete, S is a finitely

generated R-module as claimed.
It follows from the claim that the natural morphism f : X → Y = Spec(R) is finite.
By Proposition 49.9 and Step 1,
f∗ ◦ (dX )2 = (dY )2 ◦ f∗ = 0.
The rings R and S have isomorphic residue fields, hence (dX )2 = 0.
Step 3: X = Spec(S) where S is a two-dimensional (noetherian) local ring containing a
field.
Let Sb be the completion of S. The natural morphism f : Y = Spec(S) b → X is flat of
relative dimension 0. By Proposition 49.23 and Step 2,
g ∗ ◦ (dX )2 = (dY )2 ◦ g ∗ = 0.
The rings Sb and S have isomorphic residue fields, hence (dX )2 = 0.
Step 4: X is an arbitrary excellent scheme over a field.
Let x and x0 be two points of X such x0 is of codimension 2 in {x}. We need to show

that the (x, x0 )-component of (dX )2 is trivial. We may assume that X = {x}. The ring
S = OX,x0 is local 2-dimensional. The natural morphism f : Y = Spec(S) → X is flat of
constant relative dimension. By Proposition 49.23 and Step 3,
f ∗ ◦ (dX )2 = (dY )2 ◦ f ∗ = 0.
The field κ(x0 ) and the residue field of S are isomorphic, therefore, the (x, x0 )-component
of (dX )2 is trivial. ¤
49.E. Boundary map. Let X be a scheme of finite type over a field and Z ⊂ X a
closed subscheme. Set U = X \ Z. For every p ≥ 0, the set X(p) is the disjoint union of
Z(p) and U(p) , hence
Cp (X) = Cp (Z) ⊕ Cp (U ).
Consider the closed embedding i : Z → X and the open immersion j : U → X. The
sequence of complexes
∗i j∗
0 → C∗ (Z) −
→ C∗ (X) −
→ C∗ (U ) → 0
is exact. This sequence is not split in general as a sequence of complexes, but it splits
canonically termwise. Let v : C∗ (U ) → C∗ (X) and w : C∗ (X) → C∗ (Z) be the canonical
inclusion and projection. Note that v and w do not commute with the differentials in
general. We have j ∗ ◦ v = id and w ◦ i∗ = id.
We define the boundary map
∂ZU : Cp (U ) → Cp−1 (Z)
by ∂ZU := w ◦ dX ◦ v.
Example 49.31. Let X = A1F , Z = {0}, and U = Gm := A1F \ {0}. Then
¡ ¢
∂ZU {t} · [U ] = [Z],
where t is the coordinate function on A1F .
Proposition 49.32. Let X be a scheme and Z ⊂ X a closed subscheme. Set U =
X \ Z. Then dZ ◦ ∂ZU = −∂ZU ◦ dU .
Proof. By the definition of ∂ = ∂ZU , we have i∗ ◦ ∂ = dX ◦ v − v ◦ dU . Hence by
Propositions 49.9 and 49.30,
i∗ ◦ dZ ◦ ∂ = dX ◦ i∗ ◦ ∂
= dX ◦ (dX ◦ v − v ◦ dU )
= − dX ◦ v ◦ dU
= (v ◦ dU − dX ◦ v) ◦ dU
= − i∗ ◦ ∂ ◦ dU .
Since i∗ is injective, we have dZ ◦ ∂ = −∂ ◦ dU . ¤
Proposition 49.33. Let a be an invertible function on X and let a0 , a00 be the restric-
tions of a on U and Z respectively. Then
¡ ¢ ¡ ¢
∂ZU α · {a0 } = ∂ZU (α) · {a00 } and ∂ZU {a0 } · α = −{a00 } · ∂ZU (α)
for every α ∈ C∗ (U ).
Proof. The homomorphisms v and w commute with the products. The statement
follows from Proposition 49.5. ¤
Let
i0 j0
Z 0 −−−→ X 0 ←−−− U 0
  
(49.34) 
gy

fy

hy
i j
Z −−−→ X ←−−− U
be a commutative diagram with i and i0 closed embeddings, j and j 0 open embeddings
and U = X \ Z, U 0 = X 0 \ Z 0 .
Proposition 49.35. Suppose that we have the diagram (49.34).
(1) If f , g and h are proper morphisms of schemes of finite type over a field then the
diagram
0
∂ U0
Cp (U 0 ) −−−
Z
→ Cp−1 (Z 0 )
 

h∗ y
 g∗
y
Z ∂U
Cp (U ) −−−→ Cp−1 (Z)
is commutative.
(2) Suppose that both squares in the diagram (49.34) are fiber squares. If f is flat of
constant relative dimension d then so are g and h and the diagram
Z ∂U
Cp (U ) −−−→ Cp−1 (Z)
 
 g ∗
h∗ y y
U 0
∂Z 0
Cp+d (U ) −−−→ Cp+d−1 (Z 0 )
0
is commutative.
Proof. (1) Consider the diagram
v0 d 0 w0
Cp (U 0 ) −−−→ Cp (X 0 ) −−−
X
→ Cp−1 (X 0 ) −−−→ Cp−1 (Z 0 )
   

h∗ y

f∗ y
f  g∗
y ∗ y
v X d w
Cp (U ) −−−→ Cp (X) −−−→ Cp−1 (X) −−−→ Cp−1 (Z).
The left and the right squares are commutative by the local nature of the definition of the
push-forward homomorphisms. The middle square is commutative by Proposition 49.9.
The proof of (2) is similar - one uses Proposition 49.23. As both squares of the
diagram are fiber squares, for any point z ∈ Z (respectively, u ∈ U ), the fibers Zz0 and
0
Xi(z) (respectively, Uu0 and Xj(u)
0
) are naturally isomorphic. ¤
Let Z1 and Z2 be closed subschemes of a scheme X. Set
T 1 = Z1 \ Z2 , T 2 = Z2 \ Z1 , U i = X \ Zi , U = U1 ∩ U2 , Z = Z1 ∩ Z2 .
We have the following fiber product diagram of open and closed embeddings:
Z −−−→ Z2 ←−−− T2
  
  
y y y
Z1 −−−→ X ←−−− U1
x x x
  
  
T1 −−−→ U2 ←−−− U.
Denote by ∂t , ∂b , ∂l , ∂r the boundary homomorphisms for the top, bottom, left and right
triples of the diagram respectively.
Proposition 49.36. The morphism
∂l ◦ ∂b + ∂t ◦ ∂r : C∗ (U ) → C∗−2 (Z)
is homotopic to zero.
Proof. The differential of C∗ (X) relative to the decomposition
C∗ (X) = C∗ (U ) ⊕ C∗ (T1 ) ⊕ C∗ (T2 ) ⊕ C∗ (Z)
is given by the matrix  
dU ∗ ∗ ∗
 ∂b ∗ ∗ ∗ 
dX = 
 ∂r ∗ ∗ ∗ 

h ∂l ∂t dZ
where h : C∗ (U ) → C∗−1 (Z) is some morphism. The equality (dX )2 = 0 gives
h ◦ dU + dZ ◦ h + ∂l ◦ ∂b + ∂t ◦ ∂r = 0.
In other words, −h is a contracting homotopy for ∂l ◦ ∂b + ∂t ◦ ∂r . ¤
50. External products

From now on the word “scheme” means a separated scheme of finite type over a field.
Let X and Y be two schemes over F . We define the external product
Cp (X) × Cq (Y ) → Cp+q (X × Y ), (α, β) 7→ α × β
as follows: For a point v ∈ (X × Y )(p+q) , we set (α × β)v = 0 unless the point v projects
to a point x in X(p) and y in Y(q) . In the latter case
(α × β)v = lv · rF (v)/F (x) (αx ) · rF (v)/F (y) (βy ),
¡ ¢ ¡ ¢
where lv is the length of the artinian local ring of v on Spec F (x) × Spec F (y) .
50. EXTERNAL PRODUCTS 237
The external product is graded symmetric with respect to X and Y . More precisely,
if α ∈ Cp,n (X) and β ∈ Cq,m (Y ) then
(50.1) β × α = (−1)(p+n)(q+m) (α × β).
¡ ¢
For every point x ∈ X, we write Yx for Y × Spec F (x) and hx for the canonical
flat morphism Yx → Y of relative dimension
¡ ¢ 0. Note that Yx is a scheme over F (x), in
particular, C∗ (Yx ) is a module over K∗ F (x) . Denote by ix : Yx → X × Y the canonical
morphism. Let α ∈ Cp (X) and β ∈ Cq (Y ). Unfolding the definitions, we see that
X ¡ ¢
α×β = (ix )∗ αx · (hx )∗ (β) .
x∈X(p)
¡ ¢
Symmetrically, for every point y ∈ Y , we write Xy for X × Spec F (y) and ky for the
canonical flat morphism Xy → X of relative dimension¡ ¢ 0. Note that Xy is a scheme over
F (y), in particular, C∗ (Xy ) is a module over K∗ F (y) . Denote by jy : Xy → X × Y the
canonical morphism. Let α ∈ Cp (X) and β ∈ Cq (Y ). Then
X ¡ ¢
α×β = (jy )∗ (ky )∗ (α) · βy .
y∈Y(q)
Proposition 50.2. For every α ∈ C∗ (X), β ∈ C∗ (Y ), and γ ∈ C∗ (Z), we have

(α × β) × γ = α × (β × γ).
Proof. It is sufficient to show that for every point w ∈ (X × Y × Z)(p+q+r) projecting
to x ∈ X(p) , y ∈ Y(q) and z ∈ Z(r) respectively, the w-components of both sides of the
equality is equal to
rF (w)/F (x) (αx ) · rF (w)/F (y) (βy ) · rF (w)/F (z) (γz )
¡ ¢
times¡ the multiplicity
¢ ¡ that
¢ is the length of the local ring C of the point w on Spec F (x) ×
Spec F (y) × Spec F (z) . Let v ∈ (X × Y )(p+q) be the projection of w. The multiplicity
of the
¡ v-component
¢ ¡ of α¢× β is equal to the length of the local ring B of the point v on
Spec F (x) ×Spec F (y) . Clearly, C is flat over ¡ B. ¢Let Q be¡ the maximal
¢ ideal of B. The
factor ring C/QC is the local ring of w on Spec F (v) ×Spec F (z) . Then the multiplicity
of the w-component of the left hand side of the equality is equal to l(B) · l(C/QC). By
Corollary 102.2, the latter number is equal to l(C). The multiplicity of the right hand
side of the equality can be computed similarly. ¤
Proposition 50.3. For every α ∈ Cp,n (X) and β ∈ Cq,m (Y ) we have
dX×Y (α × β) = dX (α) × β + (−1)p+n α × dY (β).
¡ ¢ ¡ ¢
Proof. We may assume that α ∈ Kp+n F (x) and β ∈ Kq+m F (y) for some points
x ∈ X(p) and y ∈ Y(q) . For a point z ∈ (X × Y )(p+q−1) the z-components of all three terms
in the formula are trivial unless the projections of z to X and Y are specializations of x
and y respectively. By dimension count, ¡ z projects¢either to x or to y.
Consider the first case. We have dX (α) × β z = 0. The point z belongs to the
image of ix and the morphism ix factors as Yx → {x} × Y ,→ X × Y . The scheme Yx
is a localization of {x} × Y . By Remark 49.7 and Proposition 49.9, the z-components of

dX×Y ◦ (ix )∗ and (ix )∗ ◦ dYx are equal.
By Propositions 49.5 and 49.23, we have
£ ¤ £ ¡ ¢¤
dX×Y (α × β) z = dX×Y ◦ (ix )∗ α · (hx )∗ (β) z
£ ¡ ¢¤
= (ix )∗ ◦ dYx α · (hx )∗ (β) z
£ ¡ ¢¤
= (−1)p+n (ix )∗ α · dYx ◦ (hx )∗ (β) z
£ ¡ ¢¤
= (−1)p+n (ix )∗ α · (hx )∗ (dY β) z
£ ¤
= (−1)p+n α × dY (β) z .
¡ ¢
In the second case, symmetrically, we have α × dY (β) z = 0 and
dX×Y (α × β)z = (dX (α) × β)z . ¤
Proposition 50.4. Let f : X → X 0 and g : Y → Y 0 be morphisms. Then for every
α ∈ Cp (X) and β ∈ Cq (Y ) we have
(f × g)∗ (α × β) = f∗ (α) × g∗ (β).
Proof. Clearly, it suffices to consider the case that f is the identity of X. Let x ∈ X(p)
and let i0x : Yx0 → X × Y 0 , h0x : Yx0 → Y 0 , and gx : Yx → Yx0 be canonical morphisms. We
have
(1X × g) ◦ ix = i0x ◦ gx and g ◦ hx = h0x ◦ gx .
X ¡ ¢
(1X × g)∗ (α × β) = (1X × g)∗ ◦ (ix )∗ αx · (hx )∗ (β)
X ¡ ¢
= (i0x )∗ ◦ (gx )∗ αx · (hx )∗ (β)
X ¡ ¢
= (i0x )∗ αx · (gx )∗ (hx )∗ (β)
X ¡ ¢
= (i0x )∗ αx · (h0x )∗ g∗ (β)
= α × g∗ (β). ¤
Proposition 50.5. Let f : X 0 → X and g : Y 0 → Y be flat morphisms. Then for
every α ∈ Cp (X) and β ∈ Cq (Y ) we have
(f × g)∗ (α × β) = f ∗ (α) × g ∗ (β).
Proof. Clearly, it suffices to consider the case that f is the identity of X. Let x ∈ X(p)
and let i0x : Yx0 → X × Y 0 , h0x : Yx0 → Y 0 and gx : Yx0 → Yx be canonical morphisms. We
have
(1X × g) ◦ i0x = ix ◦ gx and g ◦ h0x = hx ◦ gx .
Note that the scheme Yx is a localization of {x} × Y . By Proposition 49.20 and Remark
49.22,
(1X × g)∗ ◦ (ix )∗ = (i0x )∗ ◦ (gx )∗ .
50. EXTERNAL PRODUCTS 239

X ¡ ¢
(1X × g)∗ (α × β) = (1X × g)∗ ◦ (ix )∗ αx · (hx )∗ (β)
X ¡ ¢
= (i0x )∗ ◦ (gx )∗ αx · (hx )∗ (β)
X ¡ ¢
= (i0x )∗ αx · (gx )∗ (hx )∗ (β)
X ¡ ¢
= (i0x )∗ αx · (h0x )∗ g ∗ (β)
= α × g ∗ (β). ¤
Corollary 50.6. Let f : X × Y → X be the projection. Then for every α ∈ C∗ (X),

we have f ∗ (α) = α × [Y ].
Proof. We apply Proposition 50.5 and Example 49.2 to f = 1X × g, where g : Y →

Spec(F ) is the structure morphism. ¤
Proposition 50.7. Let X and Y be schemes over F . Let Z ⊂ X be a closed sub-

scheme and U = X \ Z. Then for every α ∈ Cp (U ) and β ∈ Cq (Y ), we have
U ×Y
∂ZU (α) × β = ∂Z×Y (α × β).
¡ ¢
Proof. We may assume that β ∈ K∗ F (y) for some y ∈ Y . By Propositions 49.35(1)
and 50.4 we may also assume that Y = {y}. For any scheme V denote by k V : Vy → V
and j V : Vy → V × Y the canonical morphisms. Let v ∈ (Z × Y )(p+q−1) . The v-component
of both sides of the equality are trivial unless v belongs to the image of j Z . By Remark
U U ×Y
49.7, the v-component of j∗Z ◦ ∂Zyy and ∂Z×Y ◦ j∗U are equal. It follows from Propositions
49.33 and 49.35(2) that
£ ¤ £ ¡ ¢¤
∂ZU (α) × β v
= j∗Z (k Z )∗ (∂ZU α) · β v
£ ¡ U ¢¤
= j∗Z ∂Zyy (k U )∗ (α) · β v
£ U ¡ ¢¤
= j∗Z ◦ ∂Zyy (k U )∗ (α) · β v
£ U ×Y ¡ ¢¤
= ∂Z×Y ◦ j∗U (k U )∗ (α) · β v
£ U ×Y ¤
= ∂Z×Y (α × β) v . ¤
Proposition 50.8. Let X and Y be two schemes and let a be an invertible regular
function on X. Then for every α ∈ Cp (X) and β ∈ Cq (Y ) we have
¡ ¢
{a} · α × β = {a0 } · (α × β),
where a0 is the pull-back of a on X × Y .

Proof. Let ā be the pull-back of a on Xy . It follows from Propositions 49.8 and

49.16 that
¡ ¢ X ¡ ¢
{a} · α × β = (jy )∗ (ky )∗ ({a}α) · βy
X ¡ ¢
= (jy )∗ {ā}(ky )∗ (α) · βy
X ¡ ¢
= {a0 }(jy )∗ (ky )∗ (α) · βy
= {a0 } · (α × β). ¤
51. Deformation homomorphisms

In this section we construct deformation homomorphisms . We shall use them later
to define Gysin and pull-back homomorphisms. We follow Rost’s approach in [119] for
the definition of deformation homomorphisms. (Deformation homomorphisms are called
specialization homomorphism in [47].) Recall that we only consider separated schemes of
finite type over a field.
Let f : Y → X be a closed embedding, Df the deformation scheme and Cf the normal
cone of f . Recall that Df \ Cf is canonically isomorphic to Gm × X (cf. §104.E). We
define the deformation homomorphism as the composition
p∗ {t}
→ C∗+1 (Gm × X) −→ C∗+1 (Gm × X) −
∂
σf : C∗ (X) − → C∗ (Cf )
where p : Gm × X → X is the projection, the coordinate t of Gm is considered as an

invertible function on Gm × X and ∂ := ∂CGfm ×X is taken with respect to the open and
closed subsets of the deformation scheme Df .
Example 51.1. Let f = 1X for a scheme X. Then Df = A1 × X and Cf = X. We

claim that σf is the identity. Indeed, it suffices to prove that the composition
p∗ {t}
→ C∗+1 (Gm × X) −→ C∗+1 (Gm × X) −
∂
C∗ (X) − → C∗ (X)
is the identity. By Propositions 50.5, 50.7, 50.8, and Example 49.31, for every α ∈ C∗ (X),
we have
¡ ¢ ¡ ¢
∂ {t} · p∗ (α) = ∂ {t} · ([Gm ] × α)
¡ ¢
= ∂ ({t} · [Gm ]) × α
= ∂({t} · [Gm ]) × α
= {0} × α
= α.
The following statement is a consequence of Propositions 49.5, 49.23 and 49.32.
Proposition 51.2. Let f : Y → X be a closed embedding. Then σf ◦ dX = dCf ◦ σf .

51. DEFORMATION HOMOMORPHISMS 241
Consider the fiber product diagram

f0
Y 0 −−−→ X0
 
(51.3) 
gy

yh
f
Y −−−→ X
with f and f 0 closed embeddings. Then we have the fiber product diagram (cf. §104.E)
Cf 0 −−−→ Df 0 ←−−− Gm × X 0
  
  1×h
(51.4) ky ly y
Cf −−−→ Df ←−−− Gm × X.
Proposition 51.5. If the morphism h in diagram (51.3) is flat of relative dimension

d then the morphism k in (51.4) is flat of relative dimension d and the diagram
σf
Cp (X) −−−→ Cp (Cf )
 
 ∗
h∗ y yk
σf 0
Cp+d (X 0 ) −−−→ Cp+d (Cf 0 )
is commutative.
Proof. By Proposition 104.23, we have Df 0 = Df ×X X 0 , hence the morphisms l
and k in the diagram (51.4) are flat of relative dimension d. It follows from Propositions
49.16, 49.18, and 49.35(2) that the diagram
p∗ {t}
C∗+1 (Gm × X) C∗+1 (Gm × X)
∂
C∗ (X) −−−→ −−−→ −−−→ C∗ (Cf )
   
   ∗
h y
∗ (1×h)∗ y (1×h)∗ y yk
p∗ {t}
C∗+d (X 0 ) −−−→ C∗+d+1 (Gm × X 0 ) −−−→ C∗+d+1 (Gm × X 0 ) −−−→ C∗+d (Cf 0 )
∂
is commutative. ¤
Proposition 51.6. If the morphism h in (51.3) is a proper morphism then the dia-
gram
σf 0
C∗ (X 0 ) −−−→ C∗ (Cf 0 )
 

h∗ y
k
y∗
σf
C∗ (X) −−−→ C∗ (Cf )
is commutative.
Proof. The natural morphism Df 0 → Df ×X X 0 is a closed embedding by Proposition
104.23, hence the morphism l in the diagram (51.4) is proper. It follows from Propositions
49.8, 49.20, and 49.35(1) that the diagram

p∗ {t}
C∗ (X 0 ) −−−→ C∗+1 (Gm × X 0 ) −−−→ C∗+1 (Gm × X 0 ) −−−→ C∗ (Cf 0 )
∂
   

h∗ y

(1×h)∗ y

(1×h)∗ y
k
y∗
p∗ {t}
C∗ (X) −−−→ C∗+1 (Gm × X) −−−→ C∗+1 (Gm × X) −−−→ C∗ (Cf )
∂
is commutative. ¤
Corollary 51.7. Let f : Y → X be a closed embedding. Then the composition σf ◦f∗
coincides with the push-forward map C∗ (Y ) → C∗ (Cf ) for the zero section Y → Cf .
Proof. The statement follows from Proposition 51.6, applied to the fiber product
square
Y Y
° 
° 
° fy
f
Y −−−→ X
and Example 51.1. ¤
Lemma 51.8. Let f : X → A1 × W be a morphism. Suppose that the¡ composition ¢
X → A1 × W → A1 and the restriction of f on both f −1 (Gm × W ) and f −1 {0} × W is
flat. Then f is flat.
Proof. Let x ∈ X, y = f (x), and z ∈ A1 the projection of y. Set A = OA1 ,z ,
B = OA1 ×W,y , and C = OX,x . We need to show that C is flat over B. If z 6= 0, this follows
from the flatness of the restriction of f on f −1 (Gm × W ).
Suppose that z = 0. Let M be the maximal ideal ¡ of A. ¢The rings B/M B and C/M C
−1
are the local rings of y on {0}×W and of x on f {0}×W respectively. By assumption,
C/M C is flat over B/M B and C is flat over A. It follows from [100, 20G] that C is flat
over B. ¤
Lemma 51.9. Let f : U → V be a closed embedding and g : V → W a flat morphism.
Suppose that the composition
f g
q : Cf → U →
− V −
→W
is flat. Then σf ◦ g ∗ = q ∗ .
1×g
Proof. Consider the composition u : Df → A1 × V −−→ A1 × W . The restriction
of u on u−1 (Gm × W ) is isomorphic
¡ to¢ 1 × g : Gm × V → Gm × W and is therefore flat.
−1
The restriction of u on u W × {0} coincides with q and is flat by assumption. The
projection Df → A1 is also flat. It follows from Lemma 51.8 that the morphism u is flat.
Cf −−−→ Df ←−−− Gm × V
  

qy

uy

1×g y
W −−−→ A1 × W ←−−− Gm × W.
51. DEFORMATION HOMOMORPHISMS 243
By Propositions 49.16, 49.18, and 49.35(2), the following diagram is commutative:

p∗ {t}
C∗ (W ) −−−→ C∗+1 (Gm × W ) −−−→ C∗+1 (Gm × W ) −−−→
∂
C∗ (W )
   

g∗ y

1×g ∗ y

1×g ∗ y
q ∗
y
{t}
C∗+d (V ) −−−→ C∗+d+1 (Gm × V ) −−−→ C∗+d+1 (Gm × V ) −−−→ C∗+d (Cf ),
∂
where d is the relative dimension of g. As the composition in the top row of the diagram
is the identity by Example 51.1, the result follows. ¤
If f : Y → X is a regular closed embedding, we shall write Nf for the normal bundle
Cf .
Let g : Z → Y and f : Y → X be regular closed embeddings. Then f ◦ g : Z → X
is also a regular closed embedding by Proposition 104.15. The normal bundles of the
regular closed embeddings i : Nf |Z → Nf and j : Ng → Nf ◦g are canonically isomorphic;
we denote them by N (cf. §104.E).
Lemma 51.10. In the setup above, the morphisms of complexes σi ◦ σf and σj ◦ σf ◦g :
C∗ (X) → C∗ (N ) are homotopic.
Proof. Let D = Df,g be the double deformation scheme (cf. §104.F). We have the
following fiber product diagram of open and closed embeddings:
N −−−→ Di ←−−− Nf × Gm
  
  
y y y
Dj −−−→ D ←−−− Df × Gm
x x x
  
  
Gm × Nf ◦g −−−→ Gm × Df g ←−−− Gm × X × Gm .
We shall use the notation ∂t , ∂b , ∂l , ∂r for the boundary morphisms as in §49.E. For every
scheme V , denote by pV either of the projections V × Gm → V or Gm × V → V . Write p
for the projection Gm × X × Gm → X.
By Proposition 49.33 and 51.5, we have
σi ◦ σf = ∂t ◦ {s} ◦ p∗Nf ◦ σf
= ∂t ◦ {s} ◦ σf ×Gm ◦ p∗X
= ∂t ◦ {s} ◦ ∂r ◦ {t} ◦ p∗
= − ∂t ◦ ∂r ◦ {s, t} ◦ p∗ .
and similarly
σj ◦ σf g = ∂l ◦ {t} ◦ p∗Nf ◦g ◦ σf g
= ∂l ◦ {t} ◦ σGm ×f ◦g ◦ p∗X
= ∂l ◦ {t} ◦ ∂b ◦ {s} ◦ p∗
= − ∂l ◦ ∂b ◦ {t, s} ◦ p∗ .
As {s, t} = −{t, s} (cf. §49.B) and the compositions ∂t ◦ ∂r and −∂l ◦ ∂b are homotopic
by Proposition 49.36, the result follows. ¤
52. K-homology groups
Let X be a separated scheme of finite type over a field F . The complex C∗ (X) is the
coproduct of complexes C∗q (X) over all q ∈ Z. Denote the pth homology group of the
complex C∗q (X) by Ap (X, Kq ) and call it the K-homology group of X. In other words,
Ap (X, Kq ) is the homology group of the complex
a ¡ ¢ dX a ¡ ¢ dX a ¡ ¢
Kp+q+1 F (x) −→ Kp+q F (x) −→ Kp+q−1 F (x) .
dim x=p+1 dim x=p dim x=p−1
It follows from the definition that Ap (X, Kq ) = `

0 if p + q <¡0, p <¢ 0, or p > dim X.
The group Ap (X, K−p ) is a factor group of dim x=p K0 F (x) . If Z ⊂ X is a closed
subscheme, the coset of the cycle [Z] of Z in Ap (X, K−p ) (cf. Example 49.2) will be also
denoted by [Z].
If X is the disjoint union of two schemes X1 and X2 then by (49.4), we have
Ap (X, Kq ) = Ap (X1 , Kq ) ⊕ Ap (X2 , Kq ).
Example 52.1. We have
½
¡ ¢ Kq (F ) if p = 0
Ap Spec(F ), Kq =
0 otherwise.
It follows from Fact 100.5 that
½
Kq+1 (F ) if p = 1
Ap (A1
F , Kq ) =
0 otherwise.
52.A. Push-forward homomorphisms. If f : X → Y is a proper morphism of
schemes, the push-forward homomorphism f∗ : C∗q (X) → C∗q (Y ) is a morphism of com-
plexes by Proposition 49.9. This induces the push-forward homomorphism of K-homology
groups
(52.2) f∗ : Ap (X, Kq ) → Ap (Y, Kq ).
Thus, the assignment X 7→ A∗ (X, K∗ ) gives rise to a functor from the category of schemes
and proper morphisms to the category of bi-graded abelian groups and bigraded homo-
morphisms.
Example 52.3. Let f : X → Y be a closed embedding such that f is a bijection on
points. It follows from Example 49.6 that the push-forward homomorphism f∗ in (52.2)
is an isomorphism.
52.B. Pull-back homomorphism. If g : Y → X is a flat morphism of relative
dimension d, the pull-back homomorphism g ∗ : C∗q (X) → C∗+d,q−d (Y ) is a morphism
of complexes by Proposition 49.23. This induces the pull-back homomorphism of the
K-homology groups
g ∗ : Ap (X, Kq ) → Ap+d (Y, Kq−d ).
The assignment X 7→ A∗ (X, K∗ ) gives rise to a contravariant functor from the category
of schemes and flat morphisms to the category of abelian groups.
52. K-HOMOLOGY GROUPS 245
Example 52.4. If X is a variety of dimension d over F then the flat structure mor-
phism p : X → Spec(F ) of relative dimension d induces a natural pull-back homomor-
phism ¡ ¢
p∗ : Kq F = A0 Spec(F ), Kq → Ad (X, Kq−d )
giving A∗ (X, K∗ ) a structure of a K∗ (F )-module.
Example 52.5. It follows from Example 52.1 that the pull-back homomorphism
¡ ¢
f ∗ : Ap Spec(F ), Kq → Ap+1 (A1F , Kq−1 )
given by the flat structure morphism f : A1F → Spec(F ) is an isomorphism.
52.C. Product. Let X and Y be two schemes over F . It follows from Proposition
50.3 that there is a well defined pairing
Ap (X, Kn ) ⊗ Aq (Y, Km ) → Ap+q (X × Y, Kn+m )
taking the classes of cycles α and β to the class of the external product α × β (cf. §50).
52.D. Localization. Let X be a scheme and Z ⊂ X a closed subscheme. Set U :=
X \ Z and consider the closed embedding i : Z → X and the open immersion j : U → X.
The exact sequence of complexes
∗ i j∗
0 → C∗ (Z) −
→ C∗ (X) −
→ C∗ (U ) → 0
induces the long localization exact sequence of K-homology groups
δ i∗ j∗ δ i∗
(52.6) ... −
→ Ap (Z, Kq ) −
→ Ap (X, Kq ) −
→ Ap (U, Kq ) −
→ Ap−1 (Z, Kq ) −
→ ...
The map δ is called the connecting homomorphism. It is induced by the boundary map
of complexes ∂ZU : C∗ (U ) → C∗−1 (Z) (cf. Proposition 49.32).
52.E. Deformation. Let f : Y → X be a closed embedding. It follows from Propo-
sition 51.2 that the deformation homomorphism σf of complexes induces the deformation
homomorphism of homology groups
σf : Ap (X, Kq ) → Ap (Cf , Kq ),
where Cf is the normal cone of f .
Proposition 52.7. Let Z be a closed equidimensional
¡ ¢ ¡ subscheme
¢ of a scheme X and
g : f −1 (Z) → Z the restriction of f . Then σf [Z] = h∗ [Cg ] , where h : Cg → Cf is the
closed embedding of cones.
Proof. Let i : Z → X be the closed embedding and q : Z → Spec(F ), r : Cf →
Spec(F ) the structure morphisms. Consider the diagram
¡ ¢ q∗ i∗
A0 Spec(F ), K0 −−−→ Ad (Z, K−d ) −−− → Ad (X, K−d )
°  
°  
° σg y σ f y
¡ ¢ r∗ h∗
A0 Spec(F ), K0 −−−→ Ad (Cg , K−d ) −−− → Ad (Cf , K−d ),
where d = dim Z. The left square ¡is commutative
¢ by Lemma 51.9 and the¡ right ¢ one by
∗ ∗
Proposition 51.6. Consequently, σf [Z] = σf ◦ i∗ ◦ q (1) = h∗ ◦ r (1) = h∗ [Cg ] . ¤
52.F. Continuity. Let X be a variety of dimension n and f : Y → X a dominant

morphism. Let x denote the generic point of X and Yx the generic fiber of f . For every
nonempty open subscheme U ⊂ X, the natural flat morphism gU : Yx → f −1 (U ) is of
relative dimension −n. Hence we have the pull-back homomorphism
¡ ¢
(52.8) gU∗ : C∗ f −1 (U ) → C∗−n (Yx ).
The following proposition is a straightforward consequence of definition of the complexes
C∗ .
Proposition 52.9. The pull-back homomorphism gU∗ of (52.8) induces isomorphisms
¡ ¢ ∼ ¡ ¢ ∼
colim Cp f −1 (U ) → Cp−n (Yx ) and colim Ap f −1 (U ), Kq → Ap−n (Yx , Kq+n )
for all p and q, where the colimits are taken over all nonempty open subschemes U of X.
52.G. Homotopy invariance. Let g : Y → X be a morphism of schemes over ¡ F¢.
−1
Recall that for every x ∈ X, we denote by Yx the fiber scheme g (x) = Y ×X Spec F (x)
over the field F (x).
Proposition 52.10. Let g : Y → X be a flat morphism of relative dimension d.
Suppose that for every x ∈ X, the pull-back homomorphism
¡ ¡ ¢ ¢
Ap Spec F (x) , Kq → Ap+d (Yx , Kq−d )
is an isomorphism for every p and q. Then the pull-back homomorphism
g ∗ : Ap (X, Kq ) → Ap+d (Y, Kq−d )
is an isomorphism for every p and q.
Proof. Step 1: X is a variety.
We induct on n = dim X. The case n = 0 is obvious. In general, let U ⊂ X be a nonempty
open subset and Z = X \ U with the structure of a reduced scheme. Set V = g −1 (U ) and
T = g −1 (Z). We have closed embeddings i : Z → X, k : T → Y and open immersions
j : U → X, l : V → Y . By induction, the pull-back homomorphism (g|T )∗ in the diagram
δ i
∗ j∗ δ
Ap+1 (U, Kq ) −−−→ Ap (Z, Kq ) −−−→ Ap (X, Kq ) −−−→ Ap (U, Kq ) −−−→ Ap−1 (Z, Kq
    

(g|V )∗ y

(g|T )∗ y

g∗ y

(g|V )∗ y

(g|T )∗ y
δ ∗ k l∗ δ
Ap+d+1 (V, Kq−d ) −−−→ Ap+d (T, Kq−d ) −−−→ Ap+d (Y, Kq−d ) −−−→ Ap+d (V, Kq−d ) −−−→ Ap+d−1 (T, Kq
is an isomorphism. The diagram is commutative by Propositions 49.18, 49.20, and
49.35(2).
Let x ∈ X be the generic point. By Proposition 52.9, the colimit of the homomor-
phisms
(g|V )∗ : Ap (U, Kq ) → Ap+d (V, Kq−d )
over all nonempty open subschemes U of X is isomorphic to the pull-back homomorphism
¡ ¡ ¢ ¢
Ap−n Spec F (x) , Kq+n → Ap−n+d (Yx , Kq+n−d ).
By assumption, it is an isomorphism. Taking the colimits of all terms of the diagram, we
conclude by the 5-lemma that g ∗ is an isomorphism.
52. K-HOMOLOGY GROUPS 247
Step 2: X is reduced.
We proceed by induction on the number m of irreducible components of X. The case m =
1 is Step 1. Let Z be a (reduced) irreducible component of X and let U = X \Z. Consider
the commutative diagram as in Step 1. By Step 1, we have (g|T )∗ is an isomorphism. The
pull-back (g|V )∗ is also an isomorphism by the induction hypothesis. By the 5-lemma, g ∗
is an isomorphism.
Step 3: X is an arbitrary scheme.
Let X 0 be the reduced scheme Xred . Consider the fiber product diagram
g0
Y 0 −−−→ X0
 

fy

yh
g
Y −−−→ X,
0
with f and h closed embeddings. By Proposition 49.20, we have g ∗ ◦ h∗ = f∗ ◦ g ∗ . It
0
follows from Example 52.3 that the maps f∗ and h∗ are isomorphisms. Finally, g ∗ is an
isomorphism by Step 2, hence we conclude that g ∗ is also an isomorphism. ¤
Corollary 52.11. The pull-back homomorphism
g ∗ : Ap (X, Kq ) → Ap+d (X × AdF , Kq−d )
induced by the projection g : X × AdF → X is an isomorphism. In particular,
½
Kq+d (F ) if p = d
Ap (A , Kq ) =
d
0 otherwise.
Proof. Example 52.5 and Proposition 52.10 yield the statement in the case d = 1.
The general case follows by induction. ¤
A morphism g : Y → X is called an affine bundle of rank d if g is flat and the fiber
of g over any point x ∈ X is isomorphic to the affine space AdF (x) . For example, a vector
bundle of rank d is an affine bundle of rank d.
The following statement is a useful criterion establish a morphism is an affine bundle.
We shall use it repeatedly in the sequel.
Lemma 52.12. A morphism Y → X over F is an affine bundle of rank d if for any
local commutative F -algebra R and any morphism Spec(R) → X over F , the fiber product
Y ×X Spec(R) is isomorphic to AdR over R.
Proof. Applying the condition to the local ring R = OX,x for each x ∈ X, we see
that f is flat and the fiber of f over x is the affine space AdF (x) . ¤
The following theorem implies that affine spaces are essentially negligible for K-
homology computation.
Theorem 52.13 (Homotopy Invariance). Let g : Y → X be an affine bundle of rank
d. Then the pull-back homomorphism
g ∗ : Ap (X, Kq ) → Ap+d (Y, Kq−d )
is an isomorphism for every p and q.
Proof. For every x ∈ X, we have Yx ' AdF (x) . Applying Corollary 52.11 to X =
¡ ¢
Spec F (x) , we see that the pull-back homomorphism
¡ ¡ ¢ ¢
Ap Spec F (x) , Kq → Ap+d (Yx , Kq−d )
is an isomorphism for every p and q. By Proposition 52.10, the map g ∗ is an isomorphism.
¤
Corollary 52.14. Let f : E → X be a vector bundle of rank d. Then the pull-back
homomorphism
f ∗ : Ap (X, K∗ ) → Ap+d (E, K∗−d )
is an isomorphism for every p.
53. Euler classes and Projective Bundle Theorem
In this section, we compute the K-homology for projective spaces and more generally
for projective bundles.
53.A. Euler class. Let p : E → X be a vector bundle of rank r. Let s : X → E
denote the zero section. Note that p is a flat morphism of relative dimension r and s is a
closed embedding. By Corollary 52.14, the pull-back homomorphism p∗ is an isomorphism.
We call the composition
e(E) = (p∗ )−1 ◦ s∗ : A∗ (X, K∗ ) → A∗−r (X, K∗+r )
the Euler class of E. Note that isomorphic vector bundles over X have equal Euler classes.
f g
Proposition 53.1. Let 0 → E 0 → − E − → E 00 → 0 be an exact sequence of vector
bundles over X. Then e(E) = e(E 00 ) ◦ e(E 0 ).
Proof. Consider the fiber product diagram
f
E 0 −−−→ E
 

p0 y

gy
s00
X −−−→ E 00 .
By Proposition 49.20, we have g ∗ ◦ s00∗ = f∗ ◦ p0 ∗ , hence
∗ ∗
e(E 00 ) ◦ e(E 0 ) =(p00 )−1 ◦ s00∗ ◦ (p0 )−1 ◦ s0∗
∗
=(p00 )−1 ◦ g ∗ −1 ◦ f∗ ◦ s0∗
−1
=(p00 ◦ g)∗ ◦ (f ◦ s0 )∗
=p∗ −1 ◦ s∗
=e(E). ¤
Corollary 53.2. The Euler classes of any two vector bundles E and E 0 over X
commute: e(E 0 ) ◦ e(E) = e(E) ◦ e(E 0 ).
Proof. By Proposition 53.1, we have
e(E 0 ) ◦ e(E) = e(E 0 ⊕ E) = e(E ⊕ E 0 ) = e(E) ◦ e(E 0 ). ¤
53. EULER CLASSES AND PROJECTIVE BUNDLE THEOREM 249
Proposition 53.3. Let f : Y → X be a morphism and E a vector bundle over X.

Then the pull-back E 0 = f ∗ (E) is a vector bundle over Y and
(1) If f is proper then e(E) ◦ f∗ = f∗ ◦ e(E 0 ).
(2) If f is flat then f ∗ ◦ e(E) = e(E 0 ) ◦ f ∗ .
Proof. We have two fiber product diagrams
g f
E 0 −−−→ E Y −−−→ X
   
 p  
qy y and jy yi
f g
Y −−−→ X E 0 −−−→ E
where p and q are the natural morphisms and i and j are the zero sections.
(1): By Proposition 49.20, we have p∗ ◦ f∗ = g∗ ◦ q ∗ . Hence
e(E) ◦ f∗ = (p∗ )−1 ◦ i∗ ◦ f∗
= (p∗ )−1 ◦ g∗ ◦ j∗
= f∗ ◦ (q ∗ )−1 ◦ j∗
= f∗ ◦ e(E 0 ).
(2): Again by Proposition 49.20, we have g ∗ ◦ i∗ = j∗ ◦ f ∗ . Hence
f ∗ ◦ e(E) = f ∗ ◦ (p∗ )−1 ◦ i∗
= (q ∗ )−1 ◦ g ∗ ◦ i∗
= (q ∗ )−1 ◦ j∗ ◦ f ∗
= e(E 0 ) ◦ f ∗ . ¤
Proposition 53.4. Let p : E → X and p0 : E 0 → X 0 be vector bundles. Then
e(E × E 0 )(α × α0 ) = e(E)(α) × e(E 0 )(α0 )
for every α ∈ A∗ (X, K∗ ) and α0 ∈ A∗ (X 0 , K∗ ).
Proof. Let s : X → E and s0 : X 0 → E 0 be zero sections. It follows from Propositions
50.4 and 50.5 that
−1
e(E × E 0 )(α × α0 ) = (p × p0 )∗ ◦ (s × s0 )∗ (α × α0 )
∗ −1
= (p∗ −1 × p0 ) ◦ (s∗ × s0∗ )(α × α0 )
¡ ¢ ¡ ∗ −1 0 0 ¢
= p∗ −1 ◦ s∗ (α) × p0 ◦ s∗ (α )
= e(E)(α) × e(E 0 )(α0 ). ¤
Proposition 53.5. The Euler class e(1) is trivial.
Proof. It suffices to proof that the push-forward homomorphism s∗ for the zero
section s : X → A1 × X is trivial.
¡ Let
¢ t be the coordinate
¡ ¢ function on A1 . We view {t}
as an element of C1 (A1 ) = K1 F (A1 ) . Clearly, dA1 {t} = div(t) = [0]. It follows from
Proposition 50.3 that for every α ∈ A∗ (X, K∗ ), one has
¡ ¢ ¡ ¢
s∗ (α) = [0] × α = dA1 {t} × α = dA1 ×X {t} × α = 0
in A∗ (A1 × X, K∗ ) ¤
53.B. K-homology of projective spaces. Let V be a vector space of dimension

d + 1 over F , X the projective space PF (V ) and Vp a subspace of V of dimension p + 1
for p = 0, . . . , d. We view P(Vp ) as a subvariety
¡ ¢of X. Let xp ∈ X be the generic point of
P(Vp ). Consider the generator 1p of K0 F (xp ) = Z viewed as a subgroup of Cp,−p (X).
We claim that the class lp of the generator 1p in Ap (X, K−p ) does not depend on the
choice of Vp .
The statement is trivial if p = d. Suppose p < d and let Vp0 be another subspace of
dimension p + 1. We may assume that Vp and Vp0 are subspaces of a subspace W ⊂ V of
dimension p + 2. Let h and h0 be linear forms on W satisfying ker(h) = Vp and ker(h0 ) =
Vp0 . View the ratio f = h/h0 as a rational function on PF (W ), so div(f ) = 1p − 1p0 .
By definition of the K-homology group Ap (X, K−p ), the classes lp and lp0 of 1p and 1p0
respectively in Ap (X, K−p ) coincide.
Let X be a scheme over F and set PdX = PdF × X. For every i = 0, . . . , d consider the
external product homomorphism
¡ ¢
A∗−i (X, K∗+i ) → A∗ PdX , K∗ , α 7→ li × α.
The following proposition computes the K-homology of the projective space PdX .
Proposition 53.6. For any scheme X, the homomorphism
d
a ¡ ¢
A∗−i (X, K∗+i ) → A∗ PdX , K∗
i=0
P P
taking αi to li × αi is an isomorphism.
Proof. We induct on d. The case d = 0 is obvious since PdX = X. If d > 0 we view
P as a closed subscheme of PdX with open complement AdX . Consider the closed and
d−1
X
open embeddings f : PX d−1
→ PdX and g : AdX → PdX . In the diagram
` `d
0 −−−→ d−1 i=0 A∗−i (X, K∗+i ) −−−→ i=0 A∗−i (X, K∗+i ) −−−→ A∗−d (X, K∗+d ) −−−→ 0
  
  
y y h y
∗
¡ ¢ f∗ ¡ d ¢ g∗ ¡ ¢
A∗ Pd−1 P −→ A∗ AdX , K∗
δ δ
. . . −−−→ X , K ∗ −− −→ A ∗ X , K ∗ −− −−−→ . . .
the bottom row is the localization exact sequence and h : AdX → X is the canonical
morphism. The left square is commutative by Proposition 50.4 and the right square by
Proposition 50.5.
Let q : PdX → X be the projection. Since h = q ◦ g, we have h∗ = g ∗ ◦ q ∗ . By Corollary
52.11, we have h∗ is an isomorphism, hence g ∗ is surjective. Therefore, all connecting
homomorphisms δ in the bottom localization exact sequence are trivial. It follows that
the map f∗ is injective, i.e., the bottom sequence of the two maps f∗ and g ∗ is short exact.
By the induction hypothesis, the left vertical homomorphism is an isomorphism. By the
5-lemma so is the middle one. ¤
53. EULER CLASSES AND PROJECTIVE BUNDLE THEOREM 251
Corollary 53.7.
½
Kp+q (F ) · lp if 0 ≤ p ≤ d
Ap (P d
F , Kq ) =
0 otherwise.
Example 53.8. Let L be the canonical line bundle over X = PdF . We claim that
e(L)(lp ) = lp−1 for every p = 1, . . . , d. First consider the case p = d. By §104.C, we
have L = Pd+1 \ {0}, where 0 = [0 : . . . 0 : 1] and the morphism f : L → X takes
[S0 : · · · : Sn : Sn+1 ] to [S0 : · · · : Sn ]. The image Z of the zero section s : X → L
is given by Sn+1 = 0. Let H ⊂ X be the hyperplane given by S0 = 0. We have
div(Sn+1 /S0 ) = [Z] − [f −1 (H)] and therefore, in Ad−1 (X, K1−d ):
¡ ¢ ¡ ¢ ¡ ¢
e(L)(ld ) = (f ∗ )−1 s∗ ([X]) = (f ∗ )−1 [Z] = (f ∗ )−1 [f −1 (H)] = [H] = ld−1 .
In the general case consider a linear closed embedding g : PpF → PdF . The pull-back
L0 := g ∗ (L) is the canonical bundle over PpF . By the first part of the proof and Proposition
53.3(1), ¡ ¢ ¡ ¢
e(L)(lp ) = e(L) g∗ (lp ) = g∗ e(L0 )(lp ) = g∗ (lp−1 ) = lp−1 .
Example 53.9. Let L0 be the tautological line bundle over X = PdF . Similar to
Example 53.8, we get e(L0 )(lp ) = −lp−1 for every p ∈ [1, d].
53.C. Projective Bundle Theorem. Let E → X be a vector bundle of rank r > 0.
Consider the associated projective bundle morphism q : P(E) → X. Note that q is a flat
morphism of relative dimension r − 1. Let L → P(E) be either the canonical or the
tautological line bundle and e the Euler class of L.
Theorem 53.10 (Projective Bundle Theorem). Let E → X be a vector bundle of rank
r > 0. Then the homomorphism
r
a r
a ¡ ¢
ϕ(E) := er−i ◦ q ∗ : A∗−i+1 (X, K∗+i−1 ) → A∗ P(E), K∗
i=1 i=1
¡ ¢
is an isomorphism. In other words, every α ∈ A∗ P(E), K∗ can be written in the form
r
X ¡ ¢
α= er−i q ∗ (αi )
i=1
for uniquely determined elements αi ∈ A∗−i+1 (X, K∗+i−1 ).
Proof. Suppose that L is the canonical line bundle; the case of the tautological
bundle is treated similarly. If E is a trivial vector bundle, we have P(E) = X × Pr−1
F . Let
L0 be the canonical line bundle over Pr−1F . It follows from Example 53.8 that
¡
r−i ∗
¢
e(L) q (α) = e(L)r−i (lr−1 × α)
= e(L0 )r−i (lr−1 ) × α
= li−1 × α.
Hence the map ϕ(E) coincides with the one in Proposition 53.6, consequently is an iso-
morphism.
In general, we induct on d = dim X. If d = 0, the vector bundle is trivial. If d > 0

choose an open subscheme U ⊂ X such that dimension of Z = X \ U is less than d and
the vector bundle E|U is trivial. In the diagram
` ` `
. . . −−−→ ri=1 A∗−i+1 (Z, K∗+i−1 ) −−−→ ri=1 A∗−i+1 (X, K∗+i−1 ) −−−→ ri=1 A∗−i+1 (U, K∗+i−1 ) −−−→
  

ϕ(E|Z )y

ϕ(E)y

ϕ(E|U )y
¡ ¢ ¡ ¢ ¡ ¢
. . . −−−→ A∗ P(E|Z ), K∗ −−−→ A∗ P(E), K∗ −−−→ A∗ P(E|U ), K∗ −−−→
with rows the localization long exact sequences, the homomorphisms ϕ(E|Z ) are isomor-
phisms by the induction hypothesis and ϕ(E|U ) are isomorphisms as E|U is trivial. The
diagram is commutative by Proposition 53.3. The statement follows by the 5-lemma. ¤
Remark 53.11. It follows from Propositions 49.18, 49.20, and 53.3 that the iso-
morphisms ϕ(E) are natural with respect to push-forward homomorphisms for proper
morphisms of the base schemes and with respect to pull-back homomorphisms for flat
morphisms.
¡ ¢
Corollary 53.12. The pull-back homomorphism q ∗ : A∗−r+1 (X, K∗+r−1 ) → A∗ P(E), K∗
is a split injection.
Proposition 53.13 (Splitting Principle). Let E → X be a vector bundle. Then there
is a flat morphism f : Y → X of constant relative dimension, say d, such that:
(1) The pull-back homomorphism f ∗ : A∗ (X, K∗ ) → A∗+d (Y, K∗−d ) is injective.
(2) The vector bundle f ∗ (E) has a filtration by subbundles with quotients line bundles.
Proof. We induct on the rank r of E. Consider the projective bundle q : P(E) → X.

The pull-back homomorphism q ∗ is injective by Corollary 53.12. The tautological line
bundle L over P(E) is a sub-bundle of the vector bundle q ∗ (E). Applying the induction
hypothesis to the factor bundle q ∗ (E)/L over P(E), we find a flat morphism g : Y → P(E)
of constant relative dimension satisfying the conditions (1) and (2). The composition
f = q ◦ g works. ¤
To prove various relations between K-homology classes, the splitting principle allows
us to assume that all the vector bundles involved have filtration by sub-bundles with line
factors.
54. Chern classes

In this section we construct Chern classes of vector bundles as operations on K-
homology.
Let E → X be a vector bundle of rank r > 0 and q : P(E) → X the associated
projective bundle. By Theorem 53.10, for every α ∈ A∗ (X, K∗ ) there exist unique αi ∈
A∗−i (X, K∗+i ), i = 0, . . . , r such that
r
r
¡¢ X∗
¡ ¢
−e q (α) = (−1)i er−i q ∗ (αi ) ,
i=1
54. CHERN CLASSES 253
where e is the Euler class of the tautological line bundle L over P(E), i.e.,
r
X ¡ ¢
(54.1) (−1)i er−i q ∗ (αi ) = 0,
i=0
where α0 = α. Thus we have obtained group homomorphisms
(54.2) ci (E) : A∗ (X, K∗ ) → A∗−i (X, K∗+i ), α 7→ αi = ci (E)(α)
for every i = 0, . . . , r. These are called the Chern classes of E. By definition, c0 (E) is
the identity. We also set ci = 0 if i > r or i < 0 and define the total Chern class of E by
c(E) := c0 (E) + c1 (E) + · · · + cr (E)
viewed as an endomorphism of A∗ (X, K∗ ). If E is the zero bundle (of rank 0), we set
c0 (E) = 1 and ci (E) = 0 if i 6= 0.
Proposition 54.3. If E is a line bundle then c1 (E) = e(E).
Proof. We have P(E) = X and L = E by Example 104.19. Therefore the equality

(54.1) reads e(E)(α) − α1 = 0, hence c1 (E)(α) = α1 = e(E)(α). ¤
Example 54.4. If L is a line bundle, then c(L) = 1 + e(L). In particular, c(1) = 1 by

Proposition 53.5.
Proposition 54.5. Let f : Y → X be a morphism and E a vector bundle over X.

Set E 0 = f ∗ (E). Then
(1) If f is proper then c(E) ◦ f∗ = f∗ ◦ c(E 0 ).
(2) If f is flat then f ∗ ◦ c(E) = c(E 0 ) ◦ f ∗ .
Proof. Let rank E = r. Consider the fiber product diagram
P(E 0 ) −−−
h
→ P(E)
 

q0 y
q
y
f
Y −−−→ X
with flat morphisms q and q 0 of constant relative dimension r − 1. Denote by e and e0 the
Euler classes of the tautological line bundle L over P(E) and L0 over P(E 0 ) respectively.
Note that L0 = h∗ (L).
(1): By Proposition 49.20, we have h∗ ◦ (q 0 )∗ = q ∗ ◦ f∗ . By the definition of Chern classes,
for every α0 ∈ A∗ (Y, K∗ ) and αi0 = ci (E 0 )(α0 ), we have:
r
X ¡ ∗ ¢
(−1)i (e0 )r−i q 0 (αi0 ) = 0.
i=0
Applying h∗ , by Propositions 49.20 and 53.3(1), we have

Ã r !
X ¡ ∗ ¢
0 = h∗ (−1)i (e0 )r−i q 0 (αi0 )
i=0
r
X ¡ ∗ ¢
= (−1)i er−i h∗ q 0 (αi0 )
i=0
Xr
¡ ¢
= (−1)i er−i q ∗ f∗ (αi0 ) .
i=0
¡ ¢
Hence ci (E) f∗ (α0 ) = f∗ (αi0 ) = f∗ ci (E 0 )(α0 ).
(2): By the definition of the Chern classes, for every α ∈ A∗ (X, K∗ ) and αi = ci (E)(α)
we have
X r
(−1)i er−i (q ∗ αi ) = 0.
i=0
∗
Applying h , by Proposition 53.3(2), we have
³X r
¡ ¢´
∗
0=h (−1)i er−i q ∗ (αi )
i=0
r
X ¡ ¢
= (−1)i (e0 )r−i h∗ q ∗ (αi )
i=0
Xr
¡ ∗ ¢
= (−1)i (e0 )r−i q 0 f ∗ (αi ) .
i=0
¡ ¢
Hence ci (E 0 ) f ∗ (α) = f ∗ (αi ) = f ∗ ci (E)(α). ¤
Proposition 54.6. Let E be a vector bundle over X possessing a filtration by sub-
bundles with factors line bundles L1 , L2 , . . . , Lr . Then for every i = 1, . . . , r, we have
¡ ¢
ci (E) = σi e(L1 ), . . . , e(Lr )
where σi is the ith elementary symmetric function, i.e.,
r
Y r
¡ ¢ Y
c(E) = 1 + e(Li ) = c(Li ).
i=1 i=1
Proof. Let q : P(E) → X be the canonical morphism and let e be the Euler class of
the tautological line bundle L over P(E). It follows from formula (54.1) and Proposition
54.5 that it suffices to prove that
r
Y ¡ ¢
e − e(q ∗ Li ) = 0
i=1
as an operation on A∗ (P(E), K∗ ). We induct on r. The case r = 1 follows from the fact

that the tautological bundle L coincides with E over P(E) = X (cf. Example 104.19).
54. CHERN CLASSES 255
In the general case, let E 0 be a subbundle of E having a filtration by subbundles with

factors line bundles L1 , L2 , . . . , Lr−1 and with E/E 0 ' Lr . Consider the natural morphism
f : U = P(E) \ P(E 0 ) → P(Lr ). Under the identification of P(Lr ) with X, the bundle Lr
is the tautological line bundle over P(Lr ). Hence f ∗ (Lr ) is isomorphic to the restriction
of L to U . In other words, L|U ' q ∗ (Lr )|U and therefore e(L|U ) = e(q ∗ (Lr )|U ). It follows
from Proposition 53.3 that for every α ∈ A∗ (P(E), K∗ ), we have
¡ ¢ ¡ ¢
e − e(q ∗ Lr ) (α)|U = e(L|U ) − e(q ∗ (Lr )|U ) (α|U ) = 0.
By the localization sequence 52.6, there is a β ∈ A∗ (P(E 0 ), K∗ ) satisfying
¡ ¢
i∗ (β) = e − e(q ∗ Lr ) (α),
where i : P(E 0 ) → P(E) is the closed embedding. Let L0 be the the tautological line
bundle over P(E 0 ) and q 0 : P(E 0 ) → X the canonical morphism. We have q 0 = q ◦ i.
By induction and Proposition 53.3,
r
Y r−1
Y
¡ ∗
¢ ¡ ¢
e − e(q Li ) (α) = e − e(q ∗ Li ) (i∗ β)
i=1 i=1
Ãr−1 !
Y¡ ¢
= i∗ e(L0 ) − e(i∗ q ∗ Li ) (β)
i=1
Ãr−1 !
Y¡ ∗ ¢
= i∗ e(L0 ) − e(q 0 Li ) (β)
i=1
= 0. ¤
f g
Proposition 54.7 (Whitney Sum Formula). Let 0 → E 0 − → E 00 → 0 be an exact
→E−
0 00
sequence of vector bundles over X. Then c(E) = c(E ) ◦ c(E ), i.e.,
X
cn (E) = ci (E 0 ) ◦ cj (E 00 )
i+j=n
for every n.
Proof. By the splitting principle (Proposition 53.13) and Proposition 54.5(2), we

may assume that E 0 and E 00 have filtrations by subbundles with quotients line bundles
L01 , . . . , L0r and L001 , . . . , L00s respectively. Consequently, E has a filtration with factors
L01 , . . . , L0r , L001 , . . . , L00s . It follows from Proposition 54.6 that
r
Y s
Y
0 00
c(E ) ◦ c(E ) = c(L0i ) ◦ c(L00i ) = c(E). ¤
i=1 j=1
The same proof as for Corollary 53.2 yields:

Corollary 54.8. The Chern classes of any two vector bundles E and E 0 over X
commute: c(E 0 ) ◦ c(E) = c(E) ◦ c(E 0 ).
By Example 54.4, we have
Corollary 54.9. If E is a vector bundle over X then c(E ⊕ 1) = c(E). In particular,

if E is a trivial vector bundle then c(E) = 1.
The Whitney Sum Formula allows us to define Chern classes not only for vector
bundles over a scheme X but also for elements of the Grothendieck group K0 (X). Note
that for a vector bundle E over X the endomorphisms ci (E) are nilpotent for i > 0;
therefore, the total Chern class c(E) is an invertible
¡ endomorphism.
¢ By the Whitney
Sum Formula, the assignment E 7→ c(E) ∈ Aut A∗ (X, K∗ ) gives rise to the total Chern
class homomorphism ¡ ¢
c : K0 (X) → Aut A∗ (X, K∗ ) .
55. Gysin and pull-back homomorphisms

In this section we consider contravariant properties of K-homology.
55.A. Gysin homomorphisms. Let f : Y → X be a regular closed embedding of
codimension r and let pf : Nf → Y be the canonical morphism (cf. §104.B). We define
the Gysin homomorphism as the composition
∗ −1
f F : A∗ (X, K∗ ) −→ A∗ (Nf , K∗ ) −−−−→ A∗−r (Y, K∗+r ).

σf (pf )
→ X be regular closed embeddings. Then (f ◦ g)F =

g f
Proposition 55.1. Let Z →
− Y −
gF ◦ f F.
Proof. The normal bundles of the regular closed embeddings i : Nf |Z → Nf and
j : Ng → Nf ◦g are canonically isomorphic, denote them by N . Consider the diagram
σf ◦g p∗f g
C∗ (X) −−−→ C∗ (Nf ◦g ) ←−−− C∗ (Z)
  °

σf y

σj y
°
°
σ (pg pj )∗
C∗ (Nf ) −−−i→ C∗ (N ) ←−−−− C∗ (Z)
x x °

p∗f 

p∗j 
°
°
σg p∗g
C∗ (Y ) −−−→ C∗ (Ng ) ←−−− C∗ (Z).
The bottom right square is commutative by Proposition 49.18. The bottom left and
upper right squares are commutative by Proposition 51.5 and Lemma 51.9 respectively.
The upper left square is commutative up to homotopy by Lemma 51.10. The statement
follows from commutativity of the diagram up to homotopy. ¤
Let
f0
Y 0 −−−→ X0
 
(55.2) 
gy

yh
f
Y −−−→ X
55. GYSIN AND PULL-BACK HOMOMORPHISMS 257
be a fiber product diagram with f and f 0 regular closed embeddings. The natural mor-
phisms i : Nf 0 → g ∗ (Nf ) of normal bundles over Y 0 is a closed embedding. The factor
bundle E = g ∗ (Nf )/Nf 0 over Y 0 is called the excess vector bundle.
Proposition 55.3 (Excess Formula). Let h be a proper morphism. Then in the
notation of diagram (55.2),
f F ◦ h∗ = g∗ ◦ e(E) ◦ f 0 .
F
Proof. Let
p : Nf → Y, p 0 : Nf 0 → Y 0 , i : Nf 0 → g ∗ (Nf ), r : g ∗ (Nf ) → Nf and t : g ∗ (Nf ) → Y 0
be the canonical morphisms. It suffices to prove that the diagram
σf 0 p0 ∗
C∗ (X 0 ) / C∗ (Nf 0 ) o C∗ (Y 0 )
MMM MMM
MMMi∗ MMMe(E)
MMM MMM
M& M
¡ ∗ ¢ t∗ M&
h∗ (ri)∗ C∗ g (Nf ) o C∗ (Y 0 )
q q
r∗ qqq g∗ qqqq
qqq qq
q
² σf ² xqqq p∗ xqqq
C∗ (X) / C∗ (Nf ) o C∗ (Y )
commutes.
The commutativity everywhere but at the top parallelogram follows by Propositions
49.20 and 51.6. Hence it suffices to show that t∗ ◦ e(E) = i∗ ◦ p0 ∗ . Consider the fiber
product diagram
i
Nf 0 −−−→ g ∗ (Nf )
 

p0 y
j
y
s
X −−−→ E,
where j is the natural morphism of vector bundles and s is the zero section. Let q : E → Y 0
be the natural morphism. It follows from the equality q ◦ j = t and Proposition 49.20
that
∗
t∗ ◦ e(E) = t∗ ◦ q ∗ −1 ◦ s∗ = j ∗ ◦ s∗ = i∗ ◦ p0 . ¤
Corollary 55.4. Suppose under the conditions of Proposition 55.3 that f and f 0 are
regular closed embeddings of the same codimension. Then f F ◦ h∗ = g∗ ◦ f 0 F .
Proof. In this case, E = 0 so e(E) is the identity. ¤
A consequence of Propositions 49.18 and 51.5 is the following:
Proposition 55.5. Suppose in the diagram (55.2) that h is a flat morphism. Then
the diagram
fF
A∗ (X, K∗ ) −−−→ A∗ (Y, K∗ )
 
 g ∗
h∗ y y
f0 F
A∗ (X 0 , K∗ ) −−−→ A∗ (Y 0 , K∗ )
is commutative.
Proposition 55.6.
¡ ¢ Let f : Y → X be a regular closed embedding of equidimensional
F
schemes. Then f [X] = [Y ].
Proof. By Example 49.13 and Proposition 52.7,
¡ ¢ ¡ ¢ ¡ ¢
f F [X] = (p∗f )−1 ◦ σf [X] = (p∗f )−1 [Nf ] = [Y ]. ¤
Lemma 55.7. Let i : U → V and g : V → W be a regular closed embedding and a flat
morphism respectively and let h = g ◦ i. If h is flat then h∗ = iF ◦ g ∗ .
Proof. Let p : Ni → U be the canonical morphism. By Lemma 51.9, we have
σi ◦ g ∗ = (h ◦ p)∗ = p∗ ◦ h∗ hence iF ◦ g ∗ = (p∗ )−1 ◦ σi ◦ g ∗ = h∗ . ¤
We turn to the study of the functorial behavior of Euler and Chern classes under Gysin
homomorphisms. The next proposition is a consequence of Corollary 55.4 and Proposition
55.5 (cf. the proof of Proposition 53.3).
Proposition 55.8. Let f : Y → X be a regular closed embedding and L a line bundle
over X. Set L0 = f ∗ (L). Then f F ◦ e(L) = e(L0 ) ◦ f F .
As is in the proof of Proposition 54.5, we get
Proposition 55.9. Let f : Y → X be a regular closed embedding and E a vector
bundle over X. Set E 0 = f ∗ (E). Then f F ◦ c(E) = c(E 0 ) ◦ f F .
Proposition 55.10. Let f : Y → X be a regular closed embedding. Then f F ◦ f∗ =
e(Nf ).
Proof. Let p : Nf → Y and s : Y → Nf be the canonical morphism and the zero
section of the normal bundle respectively. By Corollary 51.7,
f F ◦ f∗ = (p∗ )−1 ◦ σf ◦ f∗ = (p∗ )−1 ◦ s∗ = e(Nf ). ¤
Proposition 55.11. Let f : Y → X be a closed embedding given by a sheaf of locally
principal ideals I ⊂ OX . Let f 0 : Y 0 → X be the closed embedding given by the sheaf of
ideals I n for some n > 0 and g : Y → Y 0 the canonical morphism. Then
F
f 0 = n(g∗ ◦ f F ).
Proof. We define a natural finite morphism h : Df → Df 0 of deformation schemes
as follows: We may assume that X is affine, X = Spec(A), and Y = Spec(A/I). Then
Df = Spec(A)e and Df 0 = Spec(A
e0 ) where
a a
e=
A I −k tk , e0 =
A I −kn (t0 )k
k∈Z k∈Z
(cf. §104.E).
The morphism h is induced by the ring homomorphism Ae0 → Ae taking a component
−kn 0 k
I t identically to I t . In particular, the image of t is equal to tn .
−kn kn 0
The morphism h yields a commutative diagram (cf. §104.E)

Nf −−−→ Df ←−−− Gm × X
  

ry
 
hy qy
Nf 0 −−−→ Df 0 ←−−− Gm × X,
where q is the identity on X and the nth power morphism on Gm . Let ∂ (respectively, ∂ 0 )
be the boundary map with respect to the top row (respectively, the bottom row) of the
diagram. It follows from Proposition 49.35(1) that
(55.12) r ∗ ◦ ∂ = ∂ 0 ◦ q∗ .
For any α ∈ C∗ (X) we have
¡ ¢
(55.13) q∗ {t} · ([Gm ] × α) = {±t0 } · ([Gm ] × α)
since the norm of t in the field extension F (t)/F (t0 ) is equal to ±t0 . By (55.12) and
(55.13), we have
¡ ¢
(r∗ ◦ σf )(α) = (r∗ ◦ ∂) {t} · ([Gm ] × α)
¡ ¢
= (∂ 0 ◦ q∗ ) {t} · ([Gm ] × α)
¡ ¢
= ∂ 0 {±t0 } · ([Gm ] × α)
= σf 0 (α),
hence
(55.14) r∗ ◦ σf = σf 0 .
The morphism p factors into the composition of morphisms in the first row of the
commutative diagram
i j
Nf −−−→ (Nf )⊗n −−−→ Nf 0
  

py

sy py
0

g
Y Y −−−→ Y 0
of vector bundles. As the morphism i is finite flat of degree n, the composition i∗ ◦ i∗

is multiplication by n. The right square of the diagram is a fiber square. Hence by
Proposition 49.20, we have
∗
r∗ ◦ p∗ = j∗ ◦ i∗ ◦ i∗ ◦ s∗ = n(j∗ ◦ s∗ ) = n(p0 ◦ g∗ ).
It follows from (55.14) that
F ¡ ¢
f 0 = (p0 )−1 ◦ σf 0 = (p0 )−1 ◦ r∗ ◦ σf = n g∗ ◦ (p∗ )−1 ◦ σf = n(g∗ ◦ f F )
∗ ∗
as needed. ¤
55.B. The pull-back homomorphisms. Let f : Y → X be a morphism of equidi-

mensional schemes with X smooth. By Corollary 104.14, the morphism
i = (1Y , f ) : Y → Y × X
is a regular closed embedding of codimension dX = dim X with the normal bundle Ni =

f ∗ (TX ), where TX is the tangent bundle of X (cf. Corollary 104.14). The projection
p : Y × X → X is a flat morphism of relative dimension dY . Set d = dX − dY . We define
the pull-back homomorphism
(55.15) f ∗ : A∗ (X, K∗ ) → A∗−d (Y, K∗+d )
as the composition iF ◦ p∗ .
We use the same notation for the pull-back homomorphism just defined and the flat
pull-back. The following proposition justifies this notation.
Proposition 55.16. Let f : Y → X be a flat morphism of equidimensional schemes

and let X be smooth. Then the pull-back f ∗ in (55.15) coincides with the flat pull-back
homomorphism.
Proof. This follows by applying Lemma 55.7 to the closed embedding i = (1Y , f ) :
Y → Y × X and to the projection g : Y × X → X. ¤
We have the following two propositions about the compositions of the pull-back maps.
g f
Proposition 55.17. Let Z −→Y → − X be morphisms of equidimensional schemes with
X smooth and g flat. Then (f ◦ g)∗ = g ∗ ◦ f ∗ .

g
Z −−−→ Y
 

iZ y
i
yY
h
Z × X −−−→ Y × X,
where iY = (1Y , f ), iZ = (1Z , f g), h = (g, 1X ) and two projections pY : Y × X → X
and pZ : Z × X → X. We have pZ = pY ◦ h. By Propositions 49.18 and 55.5,
(f ◦ g)∗ = iF ∗
Z ◦ pZ
= iF ∗ ∗
Z ◦ h ◦ pY
= g ∗ ◦ iF ∗
Y ◦ pY
= g∗ ◦ f ∗. ¤
g f
Proposition 55.18. Let Z −
→Y → − X be morphisms of equidimensional schemes with
both Y and X smooth. Then (f ◦ g)∗ = g ∗ ◦ f ∗ .
Proof. Consider the commutative diagram

(1Z ,g)
Z NNN / Z ×Y /Y
NNN
NNN
N h (1Y ,f )
k NNN' ² ²
Z× / Y ×X
7 Y ×NX
(1Z ,f g)
pp NNN
l ppp NNN
pp p NNN
² ppp NNN ²
'/
Z ×X X
where
¡ ¢ ¡ ¢
k = (1Z , g, f g), h(z, y) = z, y, f (y) , l(z, x) = z, g(z), x ,
and all unmarked arrows are the projections. Applying the Gysin homomorphisms or the
flat pull-backs for all arrows in the diagram, we get a diagram of homomorphisms of the
K-homology groups that is commutative by Propositions 49.18, 55.1, 55.5, and Lemma
55.7. ¤
We next show that the pull-back homomorphism for a regular closed embedding co-
incides with the Gysin homomorphism:
Proposition 55.19. Let f : Y → X be a regular closed embedding of equidimensional
schemes with X smooth. Then f ∗ = f F .
Proof. The commutative diagram
d h
Y FF / Y × Y / Y ×X
FF
FF
F p q
1Y FFF ²
# ² f
Y / X,
with d the diagonal embedding and h = 1Y × f gives rise to the diagram

fF
A∗ (X, K∗ ) / A∗ (Y, K∗ )
QQQ
QQQ1
q ∗ p ∗ QQQ
QQQ
² ² (
hF / dF / A∗ (Y, K∗ ).
A∗ (Y × X, K∗ ) A∗ (Y × Y, K∗ )
The square is commutative by Proposition 55.5 and the triangle by Lemma 55.7. Let
g = h ◦ d. Then
f ∗ = g F ◦ q ∗ = dF ◦ hF ◦ q ∗ = f F . ¤
The pull-back homomorphisms commute with external products:
Proposition 55.20. Let f : X 0 → X and g : Y 0 → Y be morphisms of equidimen-
sional schemes with X and Y smooth. Then for every α ∈ C∗ (X) and β ∈ C∗ (Y ), we
have
(f × g)∗ (α × β) = f ∗ (α) × g ∗ (β).
Proof. It suffices to do the case with g = 1Y and by Proposition 50.5 with f a

regular closed embedding. Denote by qX : Gm × X → X and pf : Nf → X 0 the canonical
morphisms. Note that Nf ×1Y = Nf × Y . Consider the diagram
q∗ {t} p∗f
C∗ (Gm × X) C∗ (Gm × X)
∂
C∗ (X) X
−−−→ −−−→ −−−→ C∗ (Nf ) ←−−− C∗ (X 0 )
    
    
y y y y y
∗
qX×Y {t} (pf ×1Y )∗
C∗ (X × Y ) −−−→ C∗ (Gm × X × Y ) −−−→ C∗ (Gm × X × Y ) −−−→ C∗ (Nf ×1Y ) ←−−−−− C∗ (X 0 × Y )
∂
where all vertical homomorphisms are given by the external product with β. The com-
mutativity of all squares follow from Propositions 50.5, 50.7, and 50.8. ¤
Proposition 55.21.
¡ ¢Let f : Y → X be a morphism of equidimensional schemes with
∗
X smooth. Then f [X] = [Y ].
Proof. Let i = (1Y , f ) : Y → Y × X be the graph of f and let p : Y × X → X be
the projection. It follows from Corollary 50.6 and Proposition 55.6 that
¡ ¢ ¡ ¢ ¡ ¢
f ∗ [X] = iF ◦ p∗ [X] = iF [Y × X] = [Y ]. ¤
The following statement on the commutativity of the pull-back maps and the Chern
classes is a consequence of Propositions 54.5(2) and 55.9.
Proposition 55.22. Let f : Y → X be a morphism of equidimensional schemes with
X smooth and E a vector bundle over X. Set E 0 = f ∗ (E). Then f ∗ ◦ c(E) = c(E 0 ) ◦ f ∗ .
56. K-cohomology ring of smooth schemes
We now consider the case that our scheme X is smooth. This allows us to introduce the
K-cohomology groups A∗ (X, K∗ ) which we do as follows: If X is irreducible of dimension
d, we set
Ap (X, Kq ) := Ad−p (X, Kq−d ).
In the general case, let X1 , X2 , . . . , Xs be (disjoint) irreducible components of X. We set
as
p
A (X, Kq ) := Ap (Xi , Kq ).
i=1
In particular, if X is an equidimensional smooth scheme of dimension d, then Ap (X, Kq ) =

Ad−p (X, Kq−d ).
Let f : Y → X be a morphism of smooth schemes. We define the pull-back homo-
morphism
f ∗ : Ap (X, Kq ) → Ap (Y, Kq )
as follows: If X and Y are both irreducible and of dimension dX and dY respectively, we
define f ∗ as in §55.B:
f∗
f ∗ : Ap (X, Kq ) = AdX −p (X, Kq−dX ) −→ AdY −p (X, Kq−dY ) = Ap (Y, Kq ).
If just Y is irreducible, we have f (Y ) ⊂ Xi for an irreducible component Xi of X. We
define the pull-back as the composition
f∗
Ap (X, Kq ) → Ap (Xi , Kq ) −→ Ap (Y, Kq ),
56. K-COHOMOLOGY RING OF SMOOTH SCHEMES 263
where the first map is the canonical projection. Finally, in the general case, we define
f ∗ as the direct sum of the homomorphisms Ap (X, Kq ) → Ap (Yj , Kq ) over all irreducible
components Yj of Y .
g f
It follows from Proposition 55.18 that if Z −
→ Y −
→ X are morphisms of smooth
∗ ∗ ∗
schemes then (f ◦ g) = g ◦ f .
Let X be a smooth scheme. Denote by
d = dX : X → X × X
the diagonal closed embedding. The composition
0 × 0 d∗ 0
(56.1) → Ap+p (X × X, Kq+q0 ) −
Ap (X, Kq ) ⊗ Ap (X, Kq0 ) − → Ap+p (X, Kq+q0 )
defines a product on A∗ (X, K∗ ).
`
Remark 56.2. If X = X1 X2 then A∗ (X, K∗ ) = A∗ (X1 , K∗ ) ⊕ A∗ (X2 , K∗ ). Since
the image of the diagonal morphism dX does not intersect X1 × X2 , the product of two
classes from A∗ (X1 , K∗ ) and A∗ (X2 , K∗ ) is zero.
Proposition 56.3. The product in (56.1) is associative.
Proof. Let α, β, γ ∈ A∗ (X, K∗ ). By Proposition 55.20, we have
¡ ¢
(α × β) × γ =d∗ d∗ (α × β) × γ
= d∗ ◦ (d × 1X )∗ (α × β × γ)
¡ ¢∗
= (d × 1X ) ◦ d (α × β × γ)
= c∗ (α × β × γ),
where c : X → X × X × X is the diagonal embedding. Similarly, α × (β × γ) =

c∗ (α × β × γ). ¤
Proposition 56.4. For every smooth scheme X, the product in A∗ (X, K∗ ) is bigraded
0
commutative, i.e., if α ∈ Ap (X, Kq ) and α0 ∈ Ap (X, Kq0 ) then
0 0
α · α0 = (−1)(p+q)(p +q ) α0 · α.
Proof. It follows from (50.1) that
0 0 0 0
α · α0 = d∗ (α × α0 ) = (−1)(p+q)(p +q ) d∗ (α0 × α) = (−1)(p+q)(p +q ) α0 · α. ¤
Let X bePa smooth scheme and let X1 , X2 , . . . be the irreducible components of X.
Then [X] = [Xi ] in A0 (X, K0 ).
Proposition 56.5. The class [X] is the identity in A∗ (X, K∗ ) under the product.
Proof. We may assume that X is irreducible. Let f : X × X → X be the first
projection. Since f ◦ d = 1X , it follows from Corollary 50.6 and Proposition 55.16 that
¡ ¢
α · [X] = d∗ α × [X] = d∗ f ∗ (α) = α. ¤
We have proven:
Theorem 56.6. Let X be a smooth scheme. Then A∗ (X, K∗ ) is a bigraded commuta-

tive associative ring with the identity [X].
Remark 56.7. If X1 , . . . , Xn are the irreducible components of a smooth scheme X,
the ring A∗ (X, K∗ ) is the product of the rings A∗ (X1 , K∗ ), . . . , A∗ (Xn , K∗ ).
Proposition 56.8. Let f : Y → X be a morphism of smooth schemes. Then
f ∗ (α · β) = f ∗ (α) · f ∗ (β)
¡ ¢
for all α, β ∈ A∗ (X, K∗ ) and f ∗ [X] = [Y ].
Proof. Since (f × f ) ◦ dY = dX ◦ f , it follows from Propositions 55.18 and 55.20 that
f ∗ (α · β) =f ∗ ◦ d∗X (α × β)
= d∗Y ◦ (f × f )∗ (α × β)
¡ ¢
= d∗Y f (α) × f (β)
= f ∗ (α) · f ∗ (β).
The second equality follows from Proposition 55.21. ¤
It follows from Proposition 56.8 that the correspondence X 7→ A∗ (X, K∗ ) gives rise to
a cofunctor from the category of smooth schemes and arbitrary morphisms to the category
of bigraded rings and homomorphisms of bigraded rings.
Proposition 56.9 (Projection Formula). Let f : Y → X be a proper morphism of
smooth schemes. Then ¡ ¢
f∗ α · f ∗ (β) = f∗ (α) · β
for every α ∈ A∗ (Y, K∗ ) and β ∈ A∗ (X, K∗ ).
Proof. Let g = (1Y × f ) ◦ dY . Then we have the fiber product diagram

g
Y −−−→ Y × X
 
(56.10) 
fy
f ×1
y X
X d
X −−− → X × X.
It follows from Propositions 49.20 and 55.20 that
¡ ¢ ¡ ¢
f∗ α · f ∗ (β) =f∗ ◦ d∗Y α × f ∗ (β)
= f∗ ◦ d∗Y ◦ (1Y × f )∗ (α × β)
= f∗ ◦ g ∗ (α × β)
= d∗X ◦ (f × 1Y )∗ (α × β)
= d∗X (f∗ (α) × β)
= f∗ (α) · β. ¤
The projection formula asserts that the push-forward homomorphism f∗ is A∗ (X, K∗ )-
linear if we view A∗ (Y, K∗ ) as a A∗ (X, K∗ )-module via f ∗ .
The following statement is an analog of the projection formula.
56. K-COHOMOLOGY RING OF SMOOTH SCHEMES 265
Proposition 56.11. Let f : Y → X be a morphism of equidimensional schemes with

X smooth. Then ¡ ¢ ¡ ¢
f∗ f ∗ (β) = f∗ [Y ] · β
for every β ∈ A∗ (X, K∗ ).
Proof. The closed embeddings g and dX in the diagram (56.10) are regular of the
same codimension (cf. Corollary 104.14). Let p : X × X → X be the second projection.
Then the composition q = p◦(f ×1X ) : Y ×X → X is also the projection. By Propositions
50.4, 50.5, 55.17, 55.21, and Corollaries 50.6, 55.4, we have
¡ ¢
f∗ f ∗ (β) =f∗ ◦ g F ◦ q ∗ (β)
= f∗ ◦ g F ◦ (f × 1X )∗ ◦ p∗ (β)
= dF ∗ ∗
X ◦ (f × 1X )∗ ◦ (f × 1X ) ◦ p (β)
¡ ¢
= dF ∗
X ◦ (f∗ × id) ◦ (f × id) [X] × β
¡ ¡ ¢ ¢
= dF ∗
X f∗ ◦ f [X] × β
¡ ¢
= dF
X f∗ ([Y ]) × β
¡ ¢
= f∗ [Y ] · β. ¤
CHAPTER X
Chow groups
In this chapter we study Chow groups as special cases of K-homology and K-cohomology
theories, so we can apply results from the previous chapter. Chow groups will remain the
main tool in the rest of the book. We also develop the theory of Segre classes that will
be used in the chapter on the Steenrod operations that follows.
57. Definition of Chow groups

Recall that a scheme is a separated scheme of finite type over a field and a variety is
an integral scheme.
57.A. Two equivalent definitions of Chow groups. Let X be a scheme over F

and let p ∈ Z. We call the group
CHp (X) := Ap (X, K−p )
the Chow group of dimension p cycles on X. By definition,
 
a ¡ ¢ dX a ¡ ¢
CHp (X) = Coker  K1 F (x) −→ K0 F (x)  .
x∈X(p+1) x∈X(p)
¡ ¢ ¡ ¢
Note that K1 F (x) = F (x)× and K0 F (x) = Z. Thus the Chow group CHp (X) is the
factor group of the free abelian group
a
Zp (X) = Z,
x∈X(p)
called the group of p-dimensional cycles on X, by the subgroup generated by the divisors
dX (f ) = div(f ) for all f ∈ F (x)× and x ∈ X(p+1) .
A point x ∈ X of dimension p gives rise to a prime cycle inPZp (X), denoted by [x].
Thus, an element of Zp (X) is a finite formal linear combination nx [x] with nx ∈ Z and
dim x = p. We will often P write {x} instead of x, so that an element of Zp (X) is a finite
formal linear combination nZ [Z] where the sum is taken over closed subvarieties Z ⊂ X
of dimension p. We will use the same notation for the classes of cycles in CHp (X). Note
that a closed subscheme W ⊂ X (not necessarily integral) defines a cycle [W ] ∈ Z(X) (cf.
Example 49.2).
Example 57.1. Let X be a scheme of dimension d. The group CHd (X) = Zd (X) is
free with basis the classes of irreducible components (generic points) of X of dimension
d.
267
268 X. CHOW GROUPS
The divisor of a function can be computed in a simpler way. Let R be a 1-dimensional

Noetherian local domain with quotient field L. We define the order homomorphism
ordR : L× → Z
by the formula ordR (r) = l(R/rR) for every nonzero r ∈ R.
Let Z be a variety over F of dimension d. For any point x ∈ Z of dimension d − 1,
the local ring OZ,x is 1-dimensional. Hence the order homomorphism
ordx = ordOZ,x : F (Z)× → Z
is well defined.
Proposition
P 57.2. Let Z be a variety over F of dimension d and f ∈ F (Z)× . Then
div(f ) = ordx (f ) · x, where the sum is taken over all points x ∈ Z of dimension d − 1.
e
Proof. Let R be the local ring OZ,x , where x is a point of dimension d − 1. Let R
denote the integral closure of R in F (Z). For every nonzero f ∈ R, the x-component of
div(f ) is equal to X
eQ /f R
l(R eQ ) · [R/Q
e : F (x)],
where the sum is taken over all maximal ideals Q of R.e Applying Lemma 102.3 to the
e
R-module M = R/fe R, e we have the x-component equals lR (R/f e R).
e Since R/R
e is an
e R)
R-module of finite length, lR (R/f e = lR (R/f R) = ordx (f ). ¤
We wish to give an equivalent definition of Chow groups. To do so we need some
preliminaries.
Let Z be a variety over F of dimension d and f : Z → P1 a dominant morphism.
Thus f is a flat morphism of relative dimension d − 1. For any rational point a ∈ P1 , the
pull-back scheme f −1 (a) is an equidimensional subscheme of Z of dimension d − 1. Note
that we can view f as a rational function on Z.
Lemma 57.3. Let f be as above. Then div(f ) = [f −1 (0)] − [f −1 (∞)] on Z.
Proof. Let x ∈ Z be a point of dimension d − 1 with the x-component of div(f )
nontrivial. Then f (x) = 0 or f (x) = ∞.
Consider the first case, so f ∈ OZ,x . By Proposition 57.2, the x-component of div(f )
is equal to ordx (f ). The local ring Of −1 (0),x coincides with OZ,x /f OZ,x , therefore, the
x-component of [f −1 (0)] is equal to
l(Of −1 (0),x ) = l(OZ,x /f OZ,x ) = ordx (f ).
Similarly (applying the above argument to the function f −1 ), we see that in the second
case the x-component of [f −1 (∞)] is equal to ordx (f −1 ) = − ordx (f ). ¤
Let X be a scheme and Z ⊂ X × P1 a closed subvariety of dimension d with Z
dominant over P1 . Hence the projection f : Z → P1 is flat of relative dimension d − 1.
For every rational point a ∈ P1 , the projection p : X × P1 → X maps the subscheme
f −1 (a) isomorphically onto a closed subscheme of X that we denote by Z(a). It follows
from Lemma 57.3 that
¡ ¢
(57.4) p∗ div(f ) = [Z(0)] − [Z(∞)].
57. DEFINITION OF CHOW GROUPS 269
In particular, the classes of [Z(0)] and [Z(∞)] coincide in CH(X).

Let Z(X; P1 ) denote the subgroup of Z(X × P1 ) generated by the classes of closed
subvarieties of X × P1 that are dominant over P1 . For any cycle β ∈ Z(X; P1 ) and any
P a ∈ P , the cycle β(a) ∈ Z(X) is1 well defined. P
1
rational point
If α =
¡ nZ [Z]
¢ ∈ Z(X), we write α × [P ] for the cycle nZ [Z × P1 ] ∈ Z(X; P1 ).
Clearly, α × [P1 ] (a) = α for any rational point a ∈ P1 .
Let α and α0 be two cycles on a scheme X. We say that α and α0 are rationally
equivalent if the classes of α and α0 are equal in CH(X).
Proposition 57.5. Two cycles α and α0 on a scheme X are rationally equivalent if
and only if there is a cycle β ∈ Z(X; P1 ) satisfying α = β(0) and α0 = β(∞).
Proof. It was shown in (57.4) that the classes of the cycles β(0) and β(∞) are equal
in CH(X). Conversely, suppose that the classes of α and α0 are equal in CH(X). By
the definition of the Chow group, there are closed subvarieties Zi ⊂ X and nonconstant
rational functions gi on Zi such that
X
α − α0 = div(gi ).
Let Vi be closure of the graph of gi in Zi × P1 ⊂ X × P1 and let fi : Vi → P1 be the induced
morphism. Since gi is nonconstant, the morphism fi is dominant and [Vi ] ∈ Z(X; P1 ).
The projection p : X × P1 → X maps Vi birationally onto Zi , hence by Proposition
49.9, ¡ ¢
div(gi ) = div p∗ (fi ) = p∗ div(fi ) = [Vi (0)] − [Vi (∞)].
P
Let β 0 = [Vi ] ∈ Z(X; P1 ). We have
α − α0 = β 0 (0) − β 0 (∞).
Consider the cycle
γ = α − β 0 (0) = α0 − β 0 (∞)
and set β 00 = γ × [P1 ] and β = β 0 + β 00 . Then β(0) = β 0 (0) + β 00 (0) = β 0 (0) + γ = α and
similarly β(∞) = α0 . ¤
It follows from the above that an equivalent definition of the Chow group CH(X) is
given as the factor group of the the group of cycles Z(X) modulo the subgroup of cycles
of the form β(0) − β(∞) for all β ∈ Z(X; P1 ).
57.B. Functorial properties of the Chow groups. We now specialize the func-
torial properties developed in the previous chapter to Chow groups.
A proper morphism f : X → Y gives rise to the push-forward homomorphism
f∗ : CHp (X) → CHp (Y ).
Example 57.6. Let X be a complete scheme over F . The push-forward homomor-

phism deg : CH(X) → CH(Spec F ) = Z induced by the structure morphism X → Spec(F )
is called the degree homomorphism. For any x ∈ X, we have
½
deg(x) = [F (x) : F ] if x is a closed point
deg([x]) =
0 otherwise.
270 X. CHOW GROUPS
A flat morphism g : Y → X of relative dimension d defines the pull-back homomor-

phism
g ∗ : CHp (X) → CHp+d (Y ).
Proposition 57.7. Let g : Y → X be a flat morphism of schemes over F of¡ relative

¢
∗
dimension
£ −1 ¤ d and W ⊂ X a closed subscheme of pure dimension k. Then g [W ] =
g (W ) in Zd+k (Y ).
Proof. Consider the fiber product diagram of natural morphisms
f p
g −1 (W ) −−−→ W −−−→ Spec(F )
 

jy

yi
g
Y −−−→ X.
By Propositions 49.18 and 49.20,
¡ ¢ £ ¤
g ∗ [W ] = g ∗ ◦ i∗ ◦ p∗ (1) = j∗ ◦ f ∗ ◦ p∗ (1) = j∗ ◦ (p ◦ f )∗ (1) = g −1 (W ) . ¤
The localization property (cf. §52.D) yields:
Proposition 57.8 (Localization Sequence). Let X be a scheme, Z ⊂ X a closed
subscheme and U = X \ Z. Let i : Z → X and j : U → X be the closed embedding and
the open immersion respectively. Then the sequence
∗i j∗
CHp (Z) −
→ CHp (X) −
→ CHp (U ) → 0
is exact.
The following proposition shows that the restriction on the generic fiber of a morphism
is surjective on Chow groups.
Proposition 57.9. Let X be a variety of dimension n and f : Y → X a dominant
morphism. Let x denote the generic point of X and Yx the generic fiber of f . Then the
pull-back homomorphism CHp (Y ) → CHp−n (Yx ) is surjective.
Proof. By the continuity property (cf. Proposition 52.9), the pull-back homomor-
phism CHp (Y ) →
¡ CHp−n ¢ (Yx ) is the colimit of surjective restriction homomorphisms
CHp (Y ) → CHp f −1 (U ) over all nonempty open subschemes U of X and therefore
is surjective. ¤
Corollary 57.10. For every variety X of dimension n and scheme Y over F , the
pull-back homomorphism CHp (X × Y ) → CHp−n (YF (X) ) is surjective.
Let X and Y be two schemes. It follows from §52.C that there is a product map of
Chow groups
CHp (X) ⊗ CHq (Y ) → CHp+q (X × Y ).
Proposition 57.11. Let Z ⊂ X and W ⊂ Y be two closed equidimensional sub-
schemes of dimensions d and e respectively. Then
[Z × W ] = [Z] × [W ] in Zd+e (X × Y ).
Proof. Let p : Z → Spec(F ) and q : W → Spec(F ) be the structure morphisms and

i : Z → X and j : W → Y the closed embeddings. By Example 49.2 and Propositions
50.4, 50.5, we have
¡ ¢ ¡ ¢
[Z × W ] = (i × j)∗ ◦ (p × q)∗ (1) = i∗ ◦ p∗ (1) × j∗ ◦ q ∗ (1) = [Z] × [W ]. ¤
Theorem 57.12 (Homotopy Invariance) (Cf. Theorem 52.13). Let g : Y → X be a
flat morphism of schemes over F of relative dimension d. Suppose that for every x ∈ X,
the fiber Yx is isomorphic to the affine space AdF (x) . Then the pull-back homomorphism
g ∗ : CHp (X) → CHp+d (Y )
is an isomorphism for every p.
Theorem 57.13 (Projective Bundle Theorem) (Cf. Theorem 53.10). Let E → X

be a vector bundle of rank r, q : P(E) → X the associated projective bundle morphism,
and e the Euler class of the canonical or tautological line bundle over P(E). Then the
homomorphism
ar r
a ¡ ¢
er−i ◦ q ∗ : CH∗−i+1 (X) → CH∗ P(E)
i=1 i=1
¡ ¢
is an isomorphism, i.e., every α ∈ CH∗ P(E) can be written in the form
r
X ¡ ¢
α= er−i q ∗ (αi )
i=1
for uniquely determined elements αi ∈ CH∗−i+1 (X).
Example 57.14. Let X = P(V ), ¡where¢V is a vector space of dimension d + 1 over

F . For every p ∈ [0, d], let lp ∈ CHp P(V ) be the class of the subscheme P(Vp ) of X,
where Vp is a subspace of V of dimension p + 1. By Corollary 53.7,
½
¡ ¢ Z · lp if 0 ≤ p ≤ d
CHp P(V ) =
0 otherwise.
Let f : Y → X be a regular closed embedding of codimension r. As usual we write
Nf for the normal bundle of f . The Gysin homomorphism
f F : CH∗ (X) → CH∗−r (Y )
is defined by the formula f F = (p∗ )−1 ◦ σf , where p : Nf → Y is the canonical morphism
and σf is the deformation homomorphism.
Let Z ⊂ X be a closed subscheme of pure dimension k and set W = f −1 (Z). The
cone Cg of the restriction g : W → Z of f is of pure dimension k. Proposition 52.7 yields:
Corollary 57.15. Under the conditions of Proposition 52.7, we have f F ¡[Z]¢ =
¡ ¢
(p∗ )−1 h∗ [Cg ] .
Lemma 57.16. Let C 0 be an irreducible component of Cg . Then C 0 is an integral cone
over a closed subvariety W 0 ⊂ W with dim W 0 ≥ k − r.
272 X. CHOW GROUPS
Proof. Let N 0 be the restriction of the normal bundle Nf on W 0 . Since C 0 is a closed

subvariety of N 0 of dimension k (cf. Example 104.3), we have
k = dim C 0 ≤ dim N 0 = dim W 0 + r. ¤
Corollary 57.17. Let V ⊂ W be an irreducible component. Then there is an irre-
ducible component of Cg that is a cone over V . In particular, dim V ≥ k − r.
Proof. Let v ∈ V be the generic point. Since the canonical morphism q : Cg → W
is surjective (that is split by the zero section), there is an irreducible component C 0 ⊂ Cg
such that v ∈ im q. Clearly, im q = V , i.e., C 0 is a cone over V . ¤
We say that the scheme Z has a proper inverse image with respect to f if every
irreducible component of W = f −1 (Z) has dimension k − r.
Proposition 57.18. Let f : Y → X be a regular closed embedding of schemes over F
of codimension r and Z ⊂ X a closed equidimensional subscheme having proper inverse
image with respectPto f . Let V1 , V2 , . . . , Vs be all the irreducible components of W =
f −1 (Z), so [W ] = ni [Vi ] for some ni > 0. Then
s
¡ ¢ X
f F [Z] = m [V ], i i
i=1
for some integers mi with 1 ≤ mi ≤ ni .

Proof. Let g : W → Z be the restriction of f and Ci the restriction of the cone Cg on
Vi . Let Ni be the restriction to Vi of the normal cone Nf . As Ni is a vector bundle of rank
r over the variety Vi of dimension k −r, the variety Ni is of dimension k. Moreover, P the Ni
are all of the irreducible components of the restriction N of Nf to W and [N ] = ni [Ni ].
The cone Cg is a closed subscheme of N of pure dimension k. Hence Ci is a closed
subscheme of Ni of pure dimension k for each i. Since Ni is a variety of dimension k, the
closed embedding of Ci into Ni is an isomorphism.
P In particular, the Ci are all of the
irreducible components of Cg , so [Cg ] = mi [Ci ] with mi = l(OCg ,xi ) and where xi ∈ Cg
is the generic point of Ci . In view of Example 49.13, we have
¡ ¢ ¡ ¢
h∗ [Ci ] = [Ni ] = p∗ [Vi ]
and by Corollary 57.15,
¡ ¢ ¡ ¢ X ¡ ¢ X
f F [Z] = (p∗ )−1 h∗ [Cg ] = (p∗ )−1 mi h∗ [Ci ] = mi [Vi ].
Finally, the closed embedding h : Cg → N induces a surjective ring homomorphism

ON,yi → OCg ,xi , where yi ∈ N is the generic point of Ni . Therefore,
1 ≤ mi = l(OCg ,xi ) ≤ l(ON,yi ) = ni . ¤
Corollary 57.19. Suppose the¡conditions
¢ P of Proposition 57.18 hold and, in addition,
the scheme W is reduced. Then f F [Z] = [Vi ], i.e., all the mi = 1.
Proof. Indeed, all ni = 1, hence all mi = 1. ¤
Let Y and Z be closed subvarieties of a smooth scheme X of codimensions p and q

respectively. We say that Y and Z intersect properly if every component of Y ∩ Z has
codimension p + q.
Applying Proposition 57.18 to the regular diagonal embedding X → X × X and the
subscheme Y × Z yields the following:
Proposition 57.20. Let Y and Z be two closed subvarieties of a smooth scheme X
that intersect
P properly. Let V1 , V2 , . . . , Vs be all irreducible components of W = Y ∩ Z and
[W ] = ni [Vi ] for some ni > 0. Then
X s
[Y ] · [Z] = mi [Vi ],
i=1
for some integers mi with 1 ≤ mi ≤ ni .
Corollary 57.21. Suppose the conditions P of Proposition 57.20 hold and in addition
the scheme W is reduced. Then [Y ] · [Z] = [Vi ], i.e., all the mi = 1.
¡ ¢
Example 57.22. Let h ∈ CH1 P(V ) be the class of a hyperplane of the ¡ projective
¢
space P(V ), where V is a vector space of dimension d + 1, and lp ∈ CHp P(V ) as in
Example 57.14. Then
h · lp = lp−1
for all p ∈ [1, d]. Indeed, h = [P(U )] and lp = [P(Vp )] where U and Vp are subspaces of V of
dimensions n and p + 1 respectively. We can choose these subspaces so that the subspace
Vp−1 = U ∩ Vp has dimension p. Then ¡ P(U¢) ∩ P(Vpp) = P(Vp−1 ) and we have equality by
Corollary
¡ 57.21.
¢ It follows that CH P(V ) = Z · h for p ∈ [0, d]. In particular, the ring
p
CH P(V ) is generated by h with the one relation hd+1 = 0.

∗
If X is smooth, we write CHp (X) for the group Ap (X, Kp ) and call it the Chow group
of codimension p classes of cycles on X. We now apply the results from §56 to this
group. The graded group CH∗ (X) has the structure of a commutative associative ring
with the identity 1X . A morphism f : Y → X of smooth schemes induces a pull-back
ring homomorphism f ∗ : CH∗ (X) → CH∗ (Y ).
57.C. Cartier divisors and the Euler class. Let D be a Cartier divisor on a
variety X of dimension d and L(D) its associated line bundle over X. Let D e denote the
associated divisor in Zd−1 (X). Recall that if D is a principal Cartier divisor given by a
nonzero rational function f on X then D e = div(f ).
¡ ¢
Lemma 57.23. In the notation above, e L(D) ([X]) = [D] e in CHd−1 (X).
Proof. Let p : L(D) → X and s : X → L(D) be the canonical morphism and the
zero section respectively. Let X = ∪ Ui be an open covering and fi rational functions on
Ui giving the Cartier divisor D. Let L(D) be the locally free sheaf of sections of L(D).
The group of sections L(D)(Ui ) consists of all rational functions f on X such that f · fi
is regular on Ui . Thus we can view fi as a section of the dual bundle L(D)∨ over Ui . The
line bundle L(D) is the spectrum of the symmetric algebra
¡ ¢⊗2 2
OX ⊕ L(D)∨ · t ⊕ L(D)∨ · t ⊕ ...
274 X. CHOW GROUPS
of the sheaf L(D)∨ . The rational functions (fi · t)/fi on p−1 (Ui ) agree on intersections so
give a well defined rational function
¡ ¢ on ∗L(D).¡ ¢ We denote this function by t.
We claim that div(t) = s∗ [X] − p [D] e . The statement is of a local nature, so we
may assume that X is affine, say X = Spec(A) and D ¡is a principal
¢ Cartier divisor given
by a rational function f on X. We have L(D) = Spec A[f t] and by Proposition 49.23,
¡ ¢ ¡ ¢ ¡ ¢ ¡
div(t) = div(f t) − div(p∗ f ) = s∗ [X] − p∗ div(f ) = s∗ [X] − p∗ [D]) e
¡ ¢ ¡ ¢ ¡ ¢
proving the claim. By the claim, the classes s∗ [X] and p∗ [D] e are equal in CHd L(D) .
¡ ¢¡ ¢ ¡ ¢
Hence, e L(D) [X] = (p∗ )−1 ◦ s∗ [X] = [D]. e ¤
¡ •¢
Example 57.24. Let C = Spec S be an integral ¡cone with ¢ S • a sheaf of graded OX -
algebras as in §104.A. Consider the cone C ⊕ 1 = Spec S • [t] . The family of functions t/s
with s ∈ S 1 on the principal open subscheme D(s) of the projective bundle P(C ⊕ 1) gives
rise to a Cartier divisor D on P(C⊕1) with L(D) the canonical line bundle.
¡ ¢¡The associated
¢
divisor D coincides with P(C). It follows from Lemma 57.23 that e L(D) [P(C ⊕ 1)] =
e
[P(C)].
Proposition 57.25. Let L and L0 be line bundles over a scheme X. Then e(L⊗L0 ) =
e(L) + e(L0 ) on CH(X).
Proof. It suffices to proof that both sides of the equality coincide on the class [Z] of
a closed subvariety Z in X. Denote by i : Z → X the closed embedding. Choose Cartier
divisors D and D0 on Z so that L|Z ' L(D) and L0 |Z ' L(D0 ). Then L|Z ⊗ L0 |Z '
L(D + D0 ). By Proposition 53.3(1),
¡ ¢ ¡ ¢
e(L ⊗ L0 ) [Z] = i∗ ◦ e(L|Z ⊗ L0 |Z ) [Z]
¡ ¢¡ ¢
= i∗ ◦ e L(D + D0 ) [Z]
= i∗ [D^ + D0 ]
e + i∗ [D
= i∗ [D] e 0]
¡ ¢ ¡ ¢
= i∗ ◦ e L(D) + i∗ ◦ e L(D0 )
¡ ¢ ¡ ¢
= e(L) [Z] + e(L0 ) [Z] . ¤
Corollary 57.26. For any line bundle L over X, we have e(L∨ ) = −e(L).
58. Segre and Chern classes

In this section, we define Segre classes and consider their relations with Chern classes.
The Segre class for a vector bundle is the inverse of the Chern class. The advantage of
Segre classes is that they can be defined for arbitrary cones (not just for vector bundles
like Chern classes). We follow the book [47] for the definition of Segre classes.
58.A. Segre classes. Let C = Spec(S • ) be a cone over X (cf. §104.A). Let q :
P(C ⊕ 1) → X be the natural morphism and L the Pcanonical line bundle over P(C ⊕ 1)
• k
(cf. §104). Let e(L) denote the total Euler class k≥0 e(L) viewed as an operation on
¡ ¢
CH P(C ⊕ 1) .
58. SEGRE AND CHERN CLASSES 275
We define the Segre homomorphism

¡ ¢
sgC : CH P(C ⊕ 1) → CH(X) by
sgC = q∗ ◦ e(L)• .
¡ ¢
The class Sg(C) := sgC [P(C ⊕ 1)] in CH(X) is known as the total Segre class of C.
Proposition 58.1. If C is a cone over X then Sg(C ⊕ 1) = Sg(C).
P
Proof. If [C] = mi [Ci ], where the Ci are the irreducible components of C then
X
[P(C ⊕ 1k )] = mi [P(Ci ⊕ 1k )]
for k ≥ 1. Therefore, we may assume that C is a variety. Let L and L0 be canonical

line bundles over P(C ⊕ 12 ) and P(C ⊕ 1) respectively. We have L0 = i∗ (L), where
i : P(C ⊕ 1) → P(C ⊕ 12 ) is the closed embedding. By Example 57.24, we have
¡ ¢
e(L) [P(C ⊕ 12 )] = [P(C ⊕ 1)],
Let q : P(C ⊕ 12 ) → X be the canonical morphism. It follows from Proposition 53.3(1)
that
Sg(C ⊕ 1) = q∗ ◦ e(L)• ([P(C ⊕ 12 )])
¡ ¢
= q∗ ◦ e(L)• i∗ ([P(C ⊕ 1)])
¡ ¢•
= q∗ i∗ ◦ e i∗ (L) ([P(C ⊕ 1)])
= (q ◦ i)∗ ◦ e(L0 )• ([P(C ⊕ 1)])
= Sg(C). ¤
Proposition 58.2. Let C be a cone over a scheme X over F and i : Z → X a closed
embedding. Let D be a closed subcone of the restriction of C on Z. Then the diagram
¡ ¢ sgD
CH P(D ⊕ 1) −−−→ CH(Z)
 

j∗ y
i
y∗
¡ ¢ sgC
CH P(C ⊕ 1) −−−→ CH(X)
is¡commutative,
¢ ¡where j : P¢(D ⊕ 1) → P(C ⊕ 1) is the closed embedding. In particular,
i∗ Sg(D) = sg P(D ⊕ 1) .
C
Proof. The canonical line bundle LD over P(D⊕ 1) is the pull-back j ∗ (LC ). It follows
from the projection formula (cf. Proposition 53.3(1)) that
sgC ◦j∗ = (qC )∗ ◦ e(LC )• ◦ j∗
¡ ¢•
= (qC )∗ ◦ j∗ ◦ e j ∗ (LC )
= i∗ ◦ (qD )∗ ◦ e(LD )•
= i∗ ◦ sgD . ¤
276 X. CHOW GROUPS
If C = E is a vector bundle over X, the projection q is a flat morphism of relative

dimension r = rank E, and we define the total Segre operation s(E) on CH(X) by
s(E) : CH(X) → CH(X), s(E) = sgE ◦ q ∗ = q∗ ◦ e(L)• ◦ q ∗ .
In particular, Sg(E) = s(E)([X]).
For every k ∈ Z denote the degree k component of the operation s(E) by sk (E), so it
is the operation
sk (E) : CHn (X) → CHn−k (X) given by
(58.3) sk (E) = q∗ ◦ e(L)k+r ◦ q ∗ .

Proposition 58.4. Let f : Y → X be a morphism of schemes over F and E a vector
bundle over X. Set E 0 = f ∗ (E). Then
(1) If f is proper then s(E) ◦ f∗ = f∗ ◦ s(E 0 ).
(2) If f is flat then f ∗ ◦ s(E) = s(E 0 ) ◦ f ∗ .
P(E 0 ) −−−
h
→ P(E)
 
 q
q0 y y
f
Y −−−→ X
0
with flat morphisms q and q of constant relative dimension r − 1 where r = rank E.
Denote by e and e0 the Euler classes of the canonical line bundle L over P(E) and L0 over
P(E 0 ) respectively. Note that L0 = h∗ (L).
s(E) ◦ f∗ = q∗ ◦ e(L)• ◦ q ∗ ◦ f∗
∗
= q∗ ◦ e(L)• ◦ h∗ ◦ q 0
∗
= q∗ ◦ h∗ ◦ e(L0 )• ◦ q 0
∗
= f∗ ◦ q∗0 ◦ e(L0 )• ◦ q 0
= f∗ ◦ s(E 0 ),
and
f ∗ ◦ s(E) = f ∗ ◦ q∗ ◦ e(L)• ◦ q ∗
= q∗0 ◦ h∗ ◦ e(L)• ◦ q ∗
= q∗0 ◦ e(L0 )• ◦ h∗ ◦ q ∗
∗
= q∗0 ◦ e(L0 )• ◦ q 0 ◦ f ∗
= s(E 0 ) ◦ f ∗ . ¤
Proposition 58.5. Let E be a vector bundle over a scheme X over F . Then
½
0 if i < 0
si (E) =
id if i = 0.
Proof. Let α ∈ CH(X). We need to prove that si (E)(α) = 0 if i < 0 and s0 (E)(α) =
α. We may assume that α = [Z], where Z ⊂ X is a closed subvariety. Let i : Z → X be
the closed embedding. By Proposition 58.4(1), we have
¡ ¢ ¡ ¢
s(E)(α) = s(E) ◦ i∗ [Z] = i∗ ◦ s(E 0 ) [Z] ,
where E 0 = i∗ (E). Hence it is sufficient to prove the statement for the vector bundle E 0
over Z and the cycle [Z]. Therefore, we may assume that X is a variety of dimension d
and α = [X] in CHd (X). Since si (E)(α) ∈ CHd−i (X), by dimension count, si (E)(α) = 0
if i < 0.
To prove the second identity, by Proposition 58.4(2), we may replace X by an open
subscheme. Therefore, we can assume that E is a trivial vector bundle, i.e., P(E) = X ×
Pr−1 . Applying Example 53.8 and Proposition 53.3(2) to the projection X × Pr−1 → Pr−1 ,
we have
¡ ¢ ¡ ¢ ¡ ¢ ¡ ¢
s0 (E) [X] = q∗ ◦ e(L)r−1 ◦ q ∗ [X] = q∗ ◦ e(L)r−1 [X] × Pr−1 = q∗ [X] × P0 = [X]. ¤
Let E → X be a vector bundle of rank r. The restriction of Chern classes defined in

§54 on Chow groups provides operations
ci (E) : CH∗ (X) → CH∗−i (X), α 7→ αi = ci (E)(α)
Example 58.6. In view of Examples
¡ d ¢ 53.8 and 57.22, the class e(L) of the canonical
line bundle L over P acts on CH P = Z[h]/(h ) by multiplication by the class h of
d d+1
a hyperplane in Pd .
¡Byd ¢Example 104.20, the class of the tangent bundle of the projectived+1space P in
d
K0 P is equal to (d + 1)[L] − 1, hence c(TPd ) is multiplication by (1 + h) .
Example 58.7. For a vector bundle E, we have ci (E ∨ ) = (−1)i ci (E). Indeed, by the
Splitting Principle 53.13, we may assume that E has a filtration by subbundles with factors
line bundles L1 , L2 , . . . , Lr . The dual bundle E ∨ then has a filtration by subbundles with
factors line bundles L∨1 , L∨2 , . . . , L∨r . As e(L∨k ) = −e(Lk ) by Corollary 57.26, it follows
from Proposition 54.6 that
¡ ¢ ¡ ¢
ci (E ∨ ) = σi e(L∨1 ), . . . , e(L∨r ) = (−1)i σi e(L1 ), . . . , e(Lr ) = (−1)i ci (E)
where σi is the ith elementary symmetric function.
Let e and ẽ be the Euler classes of the tautological and the canonical line bundles over
P(E) respectively. By Corollary 57.26, we have ẽ = −e. Therefore, the formula (54.1)
can be rewritten as
r
X
(58.8) ẽ r−i ◦ q ∗ ◦ ci (E) = 0,
i=0
where q : P(E) → X is the canonical morphism.

Proposition 58.9. Let E be a vector bundle over X. Then s(E) = c(E)−1 .
278 X. CHOW GROUPS
Proof. Applying q∗ ◦ ẽ k−1 to the equality (58.8) for the vector bundle E ⊕ 1 of rank
r + 1, we get for every k ≥ 1:
X X
0= q∗ ◦ ẽ r+k−i ◦ q ∗ ◦ ci (E ⊕ 1) = sk−i (E) ◦ ci (E ⊕ 1)
i≥0 i≥0
in view of (58.3). By Corollary 54.9, we have ci (E ⊕ 1) = ci (E). As s0 (E) = 1 and

si (E) = 0 if i < 0 by Proposition 58.5, we have s(E) ◦ c(E) = 1. ¤
Proposition 58.10. Let E → X be a vector bundle and E 0 ⊂ E a subbundle of corank
r. Then
r
X ¡ ¢
(58.11) [P(E )] =
0
ẽr−i ◦ q ∗ ◦ ci (E/E 0 ) [X]
i=0
¡ ¢
in CH P(E) .
Proof. Applying (58.8) to the factor bundle E/E 0 , we have
r
X r−i ∗
e0 ◦ q 0 ◦ ci (E/E 0 ) = 0,
i=0
where q : P(E/E ) → X is the canonical morphism and e0 is the Euler class of the
0 0
canonical line bundle over P(E/E 0 ). Applying the pull-back homomorphism with respect
to the canonical morphism P(E) \ P(E 0 ) → P(E/E 0 ), we see that the restriction of the
right hand side of the formula in (58.11) to P(E) \ P(E 0 ) is trivial. By the localization
property (cf. §52.D), the right hand side in (58.11) is equal to k[P(E 0 )] for some k ∈ Z.
To determine k, we can replace X by an open subscheme of X and assume that E
and E 0 are trivial vector bundles of rank n and n − r respectively. The right hand side in
(58.11) is then equal to
¡ ¢ ¡ ¢
ẽr ◦ q ∗ [X] = ẽr [Pn−1 × X] = [Pn−r−1 × X] = [P(E 0 )],
therefore k = 1. ¤
Proposition 58.12. Let E and E 0 be vector bundles over schemes X and X 0 respec-
tively. Then
c(E × E 0 )(α × α0 ) = c(E)(α) × c(E 0 )(α0 )
for any α ∈ CH(X) and α0 ∈ CH(X 0 ).
Proof. Let p and p0 be the projections of X × X 0 to X and X 0 respectively. We claim
that for any β ∈ CH(X) and β 0 ∈ CH(X 0 ), we have
¡ ¢
(58.13) c p∗ (E) (β × β 0 ) = c(E)(β) × β 0 ,
¡ ∗ ¢
(58.14) c p0 (E 0 ) (β × α0 ) = β × c(E 0 )(β 0 ).
To prove the claim, by Proposition 54.5, we may assume that β = [X] and β 0 = [X 0 ].
Then (58.13) and (58.14) follow from Proposition 54.5(2).
Since E × E 0 = p∗ (E) ⊕ p0 ∗ (E 0 ), by the Whitney Sum Formula 54.7 and by (58.13),

(58.14), we have
¡ ∗ ¢
c(E × E 0 )(α × α0 ) = c p∗ (E) ⊕ p0 (E 0 ) (α × α0 )
¡ ¢ ¡ ∗ ¢
= c p∗ (E) ◦ c p0 (E 0 ) (α × α0 )
¡ ¢¡ ¢
= c p∗ (E) α × c(E 0 )(α0 )
= c(E)(α) × c(E 0 )(α0 ). ¤
Proposition¡ 58.15.
¢ Let E be a vector bundle over a smooth scheme X. Then
c(E)(α) = c(E) [X] · α for every α ∈ CH(X).
Proof. Consider the vector bundle E 0 = E × X over X × X. Let d : X → X × X
be the diagonal embedding. We have E = d∗ (E 0 ). By Propositions 58.12 and 55.9,
¡ ¢¡ ¢
c(E)(α) = c d∗ (E 0 ) dF ([X] × α)
¡ ¢
= dF c(E × X) [X] × α
¡ ¢
= dF c(E)([X]) × α
¡ ¢
= c(E) [X] · α. ¤
Proposition 58.15 shows that for a vector bundle E over a smooth
¡ ¢ scheme X, the Chern
class operation c(E) is multiplication by the class β = c(E) [X] . We shall sometimes
write c(E) = β to mean that c(E) is multiplication by β.
Let f : Y → X be a morphism of schemes, i.e., X is a scheme over X. Assume that
X is a smooth variety. We shall see that CH(Y ) has a natural structure of a module over
the ring CH(X). Indeed, as we saw in §55.B, the morphism
i = (1Y , f ) : Y → Y × X
is a regular closed embedding of codimension dim X. For every α ∈ CH(Y ) and β ∈
CH(X), we set
(58.16) α · β = iF (α × β).
Proposition 58.17. Let X be a smooth variety and Y a scheme over X. Then CH(Y )
is a module over CH(X) under the product defined in (58.16). Let g : Y → Y 0 be a proper
(respectively flat) morphism of schemes over X. Then the homomorphism g∗ (respectively
g ∗ ) is CH(X)-linear.
Proof. The composition of i and the projection
¡ p :¢Y × X → Y is the identity on
Y . It follows from Lemma 55.7 that α · [X] = i α × [X] = iF ◦ p∗ (α) = 1∗Y (α) = α, i.e.,
F
the identity [X] of CH(X) acts on CH(Y ) trivially.
i
Y −−−→ Y ×X
 
 
iy yh
k
Y × X −−−→ Y × X × X,
280 X. CHOW GROUPS
where k = 1Y × dX and h = i × 1X . It follows from Corollary 55.4 that for any α ∈ CH(Y )
and β, γ ∈ CH(X), we have
¡ ¢ ¡ ¢
α · (β · γ) = iF α × (β · γ) = iF k F (α · β · γ) = iF hF (α · β · γ) = iF (α · β) × γ = (α · β) · γ.
i
Y −−−→ Y × X
 

gy
g×1
y X
i0
Y 0 −−−→ Y 0 × X.
Suppose first that the morphism g is proper. By Corollary 55.4,
F F
g∗ (α · β) = g∗ ◦ iF (α × β) = i0 (g × 1X )∗ (α × β) = i0 ◦ (g∗ (α) × β) = g∗ (α) · β
for all α ∈ CH(Y ) and β ∈ CH(X).
If g is proper, it follows from Proposition 55.5 that
F F
g ∗ (α0 · β) = g ∗ ◦ iF (α0 × β) = i0 ◦ (g × 1X )∗ (α0 × β) = i0 (g ∗ (α0 ) × β) = g ∗ (α0 ) · β
for all α0 ∈ CH(Y 0 ) and β ∈ CH(X). ¤
Proposition 58.18. Let f : Y → X be a morphism of schemes with X smooth
and g : Y → Y 0 a flat morphism. Suppose that for every point y 0 ∈ Y 0 , the pull-back
homomorphism CH(X) → CH(Yy0 ) induced by the natural morphism of the fiber Yy0 to X
is surjective. Then the homomorphism
h : CH(Y 0 ) ⊗ CH(X) → CH(Y ), α ⊗ β 7→ g ∗ (α · β)
is surjective.
Proof. The proof is similar to the one given for Proposition 52.10. Obviously we
may assume that Y 0 is reduced.
Step 1. Y 0 is a variety:
We induct on n = dim Y 0 . The case n = 0 is obvious. In general, let U 0 ⊂ Y 0 be a
nonempty open subset and Z 0 = Y 0 \ U 0 have the structure of a reduced scheme. Set
U = g −1 (U 0 ) and Z = g −1 (Z 0 ). We have closed embeddings i : Z → Y , i0 : Z 0 → Y 0 and
open immersions j : U → Y , j 0 : U 0 → Y 0 . By induction, the homomorphism hZ in the
diagram
i0 ⊗1 j ∗ ⊗1
CH(Z 0 ) ⊗ CH(X) −−
∗
−→ CH(Y 0 ) ⊗ CH(X) −−−→ CH(U 0 ) ⊗ CH(X) −−−→ 0
  
hZ y
 hY y
 hU y

i
∗ j0∗
CH(Z) −−−→ CH(Y ) −−−→ CH(U ) −−−→ 0
is surjective. The diagram is commutative by Proposition 58.17.
Let y 0 ∈ Y 0 be the generic point. By Proposition 52.9, the colimit of the homomor-
phisms
(hU )∗ : CH(U 0 ) ⊗ CH(X) → CH(U )
over all nonempty open subschemes U 0 of Y 0 is isomorphic to the pull-back homomorphism

CH(X) → CH(Yy0 ) which is surjective by assumption. Taking the colimits of all terms of
the diagram, we conclude by the 5-lemma that hY is also surjective.
Step 2. Y 0 is an arbitrary scheme:
We induct on the number m of irreducible components of Y 0 . The case m = 1 is Step
1. Let Z 0 be a (reduced) irreducible component of Y 0 and let U 0 = Y 0 \ Z 0 . Consider the
commutative diagram as in Step 1. By Step 1, the map hZ is surjective. The map hU is
also surjective by the induction hypothesis. By the 5-lemma, hY is surjective. ¤
Proposition 58.19. Let C and C 0 be cones over schemes X and X 0 respectively.
Then
Sg(C × C 0 ) = Sg(C) × Sg(C 0 ) ∈ CH(X × X 0 ).
Proof. Set C e = C ⊕ 1 and C e0 = C 0 ⊕ 1. Let L and L0 be the tautological line bundles
over P(C)
e and P(C e0 ) respectively (cf. §104.D). We view L × L0 as a vector bundle over
P(C)
e × P(Ce0 ). The canonical morphism L × L0 → C e×C e0 induces a morphism
f : P(L × L0 ) → P(C e×C e0 ).
If D is a cone, we write D◦ for the complement of the zero section in D. By §104.C, we
have L◦ = C e◦ and L0 ◦ = C
e0◦ . The open subsets C e◦ × C
e0◦ in C e×C
e0 and L◦ × L0 ◦ in L × L0
are dense. Hence f maps any irreducible component of P(L × L0 ) birationally onto an
irreducible component of P(C e×C e0 ). In particular,
¡ ¢
f∗ [P(L × L0 )] = [P(C e×C e0 )].
e be the canonical line bundle over P(C
Let L e×C e0 ). Then f ∗ (L)
e is the canonical line
bundle over P(L × L0 ). Let q : P(C e×C e0 ) → X × X 0 be the natural morphism. By
Proposition 58.1 and the Projection Formula 56.9, we have
¡ ¢
Sg(C × C 0 ) = Sg (C × C 0 ) ⊕ 1
¡ ¢
e • [P(C
= q∗ ◦ e(L) e×C e0 )]
¡ ¢
e • f∗ [P(L × L0 )]
= q∗ ◦ e(L)
¡ ¢
e • [P(L × L0 )] .
= q∗ ◦ f∗ ◦ e(f ∗ L)
The normal bundle N of the closed embedding P(C) e × P(C e0 ) → L × L0 , given by
0
the zero section, coincides with L × L . By definition of the Segre class and the Segre
operation, we have
¡ ¢ ¡ ¢
e • [P(L × L0 )] = Sg(N ) = s(N ) [P(C)
p∗ ◦ e(f ∗ L) e × P(C e0 )] ,
where p : P(L × L0 ) → P(C) e × P(C e0 ) is the natural morphism. By Propositions 58.12 and
58.9, ¡ ¢ ¡ ¢ ¡ ¢
s(N ) [P(C)e × P(C e0 )] = s(L) [P(C)]
e × s(L0 ) [P(C e0 )] .
Let g : P(C)
e → X and g 0 : P(C e0 ) → X 0 be the natural morphisms and set h = g × g 0 .
By Proposition 50.4,
¡ ¢ ¡ ¢ ¡ 0 ¢
h∗ ◦s(L) [P(C)]×s(L
e 0
)([P(Ce0 )]) = g∗ ◦s(L)([P(C)])
e × g∗ ◦s(L0 )([P(C
e0 )]) = Sg(C)×Sg(C 0 ).
282 X. CHOW GROUPS
To finish the proof it is sufficient to notice that q◦f = h◦p and therefore q∗ ◦f∗ = h∗ ◦p∗ . ¤
Exercise 58.20. (Strong Splitting Principle) Let E be a vector bundle over X. Prove
that there is a flat morphism f : Y → X such that the pull-back homomorphism f ∗ :
CH∗ (X) → CH∗ (Y ) is injective and f ∗ (E) is a direct sum of line bundles.
Exercise 58.21. Let E be a vector bundle of rank r. Prove that e(E) = cr (E).
CHAPTER XI
Steenrod operations
In this chapter we develop Steenrod operations on Chow groups modulo 2. There are
two reasons why we do not consider the operations modulo an arbitrary prime integer.
Firstly, this case is sufficient for our applications as the number 2 is the only ”critical”
prime for projective quadrics. Secondly, our approach does not immediately generalize to
the case of an arbitrary prime integer.
Unfortunately, we need to assume that the characteristic of the base field is different
from 2 in this chapter as we do not know how to define Steenrod operations in character-
istic two.
In this chapter, the word scheme means a quasi-projective scheme over a field F of
characteristic not 2. We write Ch(X) for CH(X)/2 CH(X).
Let X be a scheme. Consider the homomorphism
¡ ¢ Z(X) → Ch(X) taking the class
[Z] of a closed subvariety Z ⊂ X to j∗ Sg(TZ ) modulo 2, where Sg is the total Segre
class (cf. §58.A), TZ is the tangent cone over Z (cf. Example 104.5) and j : Z → X is
the closed embedding. We shall prove that this map factors through rational equivalence
yielding the Steenrod operation modulo 2 of X (of homological type)
SqX : Ch(X) → Ch(X)
Thus we shall have
¡ ¢
SqX ([Z]) = j∗ Sg(TZ )
modulo 2. We shall see that the operations SqX commute with the push-forward ho-
momorphisms, so they can be viewed as functors from the category of schemes to the
category of abelian groups.
For a smooth scheme X, we can then define the Steenrod operation modulo 2 of X (of
cohomological type) by the formula
SqX := c(TX ) ◦ SqX .
This formula can be viewed as a Riemann-Roch type relation between two operations.
We shall show that the operation SqX commute with pull-back homomorphisms, so can
be viewed as contravariant functors from the category of smooth schemes to the category
of abelian groups.
In this chapter, we shall also prove the standard properties of the Steenrod operations.
Steenrod operations for motivic cohomology modulo a prime integer p of a smooth
scheme X were originally constructed by Voevodsky in [144]. The reduced power opera-
tions (but not the Bockstein operation) restrict to the Chow groups of X. An “elemen-
tary” construction of the reduced power operations modulo p on Chow groups (requiring
equivariant Chow groups) was given by Brosnan in [22].
283
284 XI. STEENROD OPERATIONS
59. Definition of the Steenrod operations

Consider a cyclic group G = {1, σ} of order 2. Let Y be a G-scheme such that
p : Y → Y /G is a G-torsor over a field F (cf. §105.A). For a point z ∈ Y /G, let Yz be the
fiber of p over z. By Example 105.6, we have Yz = Spec(K), where K is an étale quadratic
algebra over F (z). Then either K is a field (and the fiber Yz has only one point y) or
K = F (z) × F (z) (and the fiber has two points y1 and y2 ). In any case, every point in Yz
is unramified (cf. Proposition 105.5(2)). It follows that for the pull-back homomorphism
p∗ : Z(Y /G) → Z(Y ), we have
½
∗ [y] if K is a field
p ([z]) =
[y1 ]+[y2 ] otherwise.
Similarly, for a point y in the fiber Yz , we have:
½
2[z] if K is a field
p∗ ([y]) =
[z] otherwise.
In particular, p∗ ◦ p∗ is multiplication by 2.
We have σ(y) = y if K is a field and σ(y1 ) = y2 otherwise. In particular,
(59.1) p∗ ◦ p∗ = 1 + σ ∗ ,
where σ ∗ : Z(Y ) → Z(Y ) is the (pull-back) isomorphism induces by σ.
The cycles [y] and [y1 ] + [y2 ] generate the group Z(Y )G of G-invariant cycles. We have
proved
Proposition 59.2. Let p : Y → Y /G be a G-torsor. Then the pull-back homomor-
phism
p∗ : Z(Y /G) → Z(Y )G
is an isomorphism.
Suppose char F 6= 2. Let X be a scheme over F . The group G acts on X 2 × A1 =
X × X × A1 by σ(x, x0 , t) = (x0 , x, −t). We have (X 2 × A1 )G = ∆(X) × {0}, where
∆ : X → X 2 is the diagonal morphism. Set
¡ ¢
(59.3) UX := (X 2 × A1 ) \ ∆(X) × {0} .
The group G acts naturally on UX . The morphism p : UX → UX /G is a G-torsor (cf.
Example 105.4).
Let α ∈ Z(X) be a cycle. The cycle α2 × A1 := α × α × A1 in Z(X 2 × A1 ) is invariant
under the above G-action as is the restriction of the cycle α2 × A1 on UX . It follows from
Proposition 59.2 that the pull-back homomorphism p∗ identifies Z(UX /G) with Z(UX )G .
2
Let αG ∈ Z(UX /G) denote the cycle satisfying p∗ (αG 2
) = (α2 × A1 )|UX .
We then have a map
2
(59.4) Z(X) → Z(UX /G), α 7→ αG .
Lemma 59.5. If α and α0 are rationally equivalent cycles in Z(X) then αG
2
and α0 2G
are rationally equivalent cycles in Z(UX /G).
59. DEFINITION OF THE STEENROD OPERATIONS 285
Proof. As in §57.A, let Z(X; P1 ) denote the subgroup of Z(X × P1 ) generated by

the classes of closed subvarieties in X × P1 dominant over P1 . Let W ⊂ X × P1 and
W 0 ⊂ X 0 × P1 be two closed subvarieties dominant over P1 . The projections W → P1
and W 0 → P1 are flat and hence so is the fiber product W ×P1 W 0 → P1 . Therefore,
every irreducible component of W ×P1 W 0 is dominant over P1 , i.e., the cycle [W ×P1 W 0 ]
belongs to Z(X × X 0 ; P1 ). By linearity, this construction extends to an external product
over P1 given by
Z(X; P1 ) × Z(X 0 ; P1 ) → Z(X × X 0 ; P1 ), (β, β 0 ) 7→ β ×P1 β 0 .
By Proposition 57.11,
£ ¤ £ ¤ £ ¤ £ ¤
(59.6) (W ×P1 W 0 )(a) = W (a) × W 0 (a) = W (a) × W 0 (a)
for any rational point a of P1 . If X 0 = X and β 0 = β, write β̃ 2 for β ×P1 β.
By Proposition 57.5, there is a cycle β ∈ Z(X; P1 ) such that α = β(0) and α0 = β(∞).
Consider the cycle β̃ 2 × [A1 ] ∈ Z(X 2 × A1 ; P1 ).
Let G act on X 2 × A1 × P1 by σ(x, x0 , t, s) = (x0 , x, −t, s). The cycle β̃ 2 × [A1 ] is
G-invariant. Since UX × P1 is a G-torsor over (UX /G) × P1 , the restriction of the cycle
β̃ 2 × [A1 ] on UX × P1 gives rise to a well defined cycle
β̃G2 ∈ Z(UX /G; P1 )
satisfying
¡ ¢
(59.7) q ∗ (β̃G2 ) = β̃ 2 × [A1 ] |UX ×P1 ,
where q : UX × P1 → (UX /G) × P1 is the canonical morphism. ¡ ¢
Let Z ⊂ (UX /G) × P1 be a closed subvariety dominant over P1 . We have p−1 Z(a) =
q −1 (Z)(a) for any rational point a of P1 , where p : UX → UX /G is the canonical morphism.
It follows from Proposition 57.7 that
¡ ¢ ¡ ¢
(59.8) p∗ γ(a) = q ∗ (γ) (a)
for every cycle γ ∈ Z(UX /G; P1 ).
P
Let β = ni [Wi ]. Then applying (59.8) to γ = β̃G2 , we see by (59.6) and (59.7) that
¡ ¢ ¡ ¢ ¡ ¢
p∗ β̃G2 (a) = q ∗ (β̃G2 ) (a) = β̃ 2 × [A1 ] |UX ×P1 (a)
X £ ¤
= ni nj Wi ×P1 Wj × A1 |UX ×P1 (a)
X £ ¤
= ni nj Wi (a) × Wj (a) × A1 |UX
¡ ¢
= p∗ β(a)2G
in Z(UX ). It follows that β̃G2 (a) = β(a)2G in Z(UX /G) since p∗ is injective on cycles. In
particular, β̃G2 (0) = β(0)2G = αG
2
and β̃G2 (∞) = β(∞)2G = α0 2G , i.e., the cycles αG
2
and α0 2G
are rationally equivalent by Proposition 57.5. ¤
By Lemma 59.5, we have a well defined map (but not a homomorphism!)
2
(59.9) vX : CH(X) → CH(UX /G), [α] 7→ [αG ].
Consider the blow up BX of X 2 × A1 along ∆(X) × {0}. The exceptional divisor is

the projective cone P(TX ⊕ 1), where TX is the tangent cone of X. The open complement
BX \ P(TX ⊕ 1) is naturally isomorphic to UX and the group G = {1, σ} acts naturally
on BX (cf. §105.B).
By Proposition 105.7, the composition
i : P(TX ⊕ 1) ,→ BX → BX /G
is a locally principal divisor with normal line bundle L⊗2 where L is the canonical line
bundle over P(TX ⊕ 1).
We define a map ¡ ¢
uX : Ch(UX /G) → Ch P(TX ⊕ 1)
as follows: Let δ ∈ Ch(UX /G). By the localization property (Proposition 57.8), there is
β ∈ Ch(BX /G) such that β|(UX /G) = δ. We set
uX (δ) = iF (β).
We claim that the result is independent of the choice of β. Indeed, if β 0 ∈ Ch(BX /G)
is another element with β 0 |(UX /G) = δ then by the localization, β 0 = β + i∗ (γ) for some
γ ∈ Ch(TX ⊕ 1). By Proposition 55.10, we have (iF ◦ i∗ )(γ) = e(L⊗2 )(γ) = 2e(L)(γ) = 0
modulo 2, hence
iF (β 0 ) = iF (β) + (iF ◦ i∗ )(γ) = iF (β)
as needed.
Let q : BX → BX /G be the projection.
¡ ¢
Lemma 59.10. The composition iF ◦ q∗ : Ch(BX ) → Ch P(TX ⊕ 1) is zero.
¡ ¢
Proof. The scheme Y := q −1 P(TX ⊕ 1) is a locally principal closed subscheme of
BX . The sheaf of ideals in OBX defining Y is the square of the sheaf of ideals of P(TX ⊕ 1)
as a subscheme of BX . Let j : Y → BX be the closed embedding and p : Y → P(TX ⊕ 1)
the natural morphism. By Corollary 55.4, we have iF ◦ q∗ = p∗ ◦ j F . It follows from
Proposition 55.11 that j F is trivial modulo 2. ¤
Proposition 59.11. For every scheme X, the map uX is a homomorphism.
P
Proof.
P 0 Let p : U X → U X /G be the projection. For any two cycles α = ni [Zi ] and
0
α = ni [Zi ] on X, we have
2
p∗ (α + α0 )2G − p∗ (αG
2
) − p∗ (αG
0
) = (1 + σ ∗ )(γ),
where X £ ¤
γ= ni n0j Zi × Zj × A1 |UX ∈ Z(UX ).
i<j
Since p ◦ p∗ = 1 + σ by (59.1) and p∗ is injective on cycles by Proposition 59.2, we have

∗ ∗
2
(α + α0 )2G − αG
2 0
− αG = p∗ (γ).
Let β, β 0 , β 00 ∈ Ch(BX /G) and δ ∈ Ch(UX ) be cycles restricting to α, α0 , α + α0 and γ
respectively satisfying
β 00 − β − β 0 = q∗ (δ).
60. PROPERTIES OF THE STEENROD OPERATIONS 287
By Lemma 59.10,
uX (α + α0 ) − uX (α) − uX (α0 ) = iF (β 00 ) − iF (β) − iF (β 0 ) = (iF ◦ q∗ )(δ) = 0. ¤
Let X be a scheme. We define the Steenrod operation of homological type as the
composition
vX uX ¡ ¢ sgTX
SqX : Ch(X) −→ Ch(UX /G) −→ Ch P(TX ⊕ 1) −−−→ Ch(X),
where sgTX is the Segre homomorphism defined in §58.A. For every integer k we write
SqX
k : Ch∗ (X) → Ch∗−k (X),
for the component of SqX decreasing dimension by k.
¡ Proposition
¢ 59.12. Let Z be a closed subvariety of a scheme X. Then SqX ([Z]) =
j∗ Sg(TZ ) , where j : Z → X is the closed embedding and Sg is the total Segre class.
2
£ ¤ £ ¤
Proof. Let α = [Z] ∈ CH(X). We have vX (α) = αG = UZ /G . Set β = BZ /G ∈
£ ¤
CH(BZ /G). By Proposition 55.6, we have iF Z (β) = P(TZ ⊕ 1) , where iZ : P(TZ ⊕ 1) →
BZ /G is the closed embedding.
iF ¡ ¢ sgTZ
Ch(BZ /G) −−− Z
→ Ch P(TZ ⊕ 1) −−−→ Ch(Z)
  

k∗ y
 
y j∗ y
iF ¡ ¢ sgTX
Ch(BX /G) −−− X
→ Ch P(TX ⊕ 1) −−−→ Ch(X)
with vertical maps the push-forward homomorphisms. The diagram is commutative by
Corollary 55.4 and Proposition 58.2. The commutativity yields
¡ ¢
SqX ([Z]) = (sgTX ◦iF
X ) k∗ (β)
= (j∗ ◦ sgTZ ◦iF

Z )(β)
¡ ¢
= (j∗ ◦ sg ) [P(TZ ⊕ 1)]
TZ
¡ ¢
= j∗ Sg(TZ ) . ¤
Remark 59.13. The maps vX , uX and sgTX commute with arbitrary field extensions
hence so do Steenrod operations. More precisely, if L/F is a field extension then the
diagram
SqX
Ch(X) −−−→ Ch(X)
 
 
y y
SqXL
Ch(XL ) −−−→ Ch(XL )
commutes.
60. Properties of the Steenrod operations
In this section, we establish the standard properties of Steenrod operations of homo-
logical type.
60.A. Formula for a smooth cycle. Let Z be a smooth closed subvariety ¡ of¢ a
scheme ¡X. By
¢ Proposition
¡ ¢58.9, the total Segre class Sg(T Z ) coincides with s(T Z ) [Z] =
−1
c(TZ ) [Z] = c(−TZ ) [Z] , where c is the total Chern class. Hence by Proposition
59.12,
¡ ¢ ¡ ¢
(60.1) SqX [Z] = j∗ ◦ c(−TZ ) [Z] ,
where j : Z → X is the closed embedding.
60.B. External products.

Theorem 60.2. Let X and Y be two schemes over a field F of characteristic not two.
Then SqX×Y (α×β) = SqX (α)×SqY (β) for any α ∈ Ch(X) and β ∈ Ch(Y ). Equivalently,
X
SqX×Y
n (α × β) = SqX Y
k (α) × Sqm (β)
k+m=n
for all n.
Proof. We may assume that α = [V ] and β = [W ] where V and W are closed
subvarieties of X and Y respectively. Let i : V → X and j : W → Y be the closed
embeddings. By Propositions 50.4, 58.19, and Corollary 104.8,
SqX×Y (α × β) = (i × j)∗ ◦ Sg(TV ×W )
= (i∗ × j∗ ) ◦ Sg(TV × TW )
¡ ¢
= (i∗ × j∗ ) ◦ Sg(TV ) × Sg(TW )
= i∗ ◦ Sg(TV ) × j∗ ◦ Sg(TW )
= SqX (α) × SqY (β). ¤
60.C. Functoriality of SqX .

Lemma 60.3. Let i : Y → X be a closed embedding. Then i∗ ◦ SqY = SqX ◦i∗ .
Proof. Let Z ⊂ Y be a closed subscheme and let j : Z → Y be the closed embedding.
By Proposition 59.12, we have
¡ ¢ ¡ ¢
i∗ ◦ SqY [Z] = i∗ ◦ j∗ ◦ Sg(TZ ) = (ij)∗ ◦ Sg(TZ ) = SqX i∗ [Z] . ¤
Lemma 60.4. Let p : Pr × X → X be the projection. Then p∗ ◦ SqP
r ×X
= SqX ◦p∗ .
Proof. The group CH(Pr × X) is generated by cycles α = [Pk × Z] for all closed
subvarieties Z ⊂ X and k ≤ r by Theorem 57.13. It follows from Lemma 60.3 that we
may assume Z = X and k = r. The statement is obvious if r = 0, so we may assume
that r > 0. Since p∗ (α) = 0, we need to prove that p∗ SqP ×X (α) = 0.
r
By Theorem 60.2, we have

r¡ ¢ ¡ ¢
SqP ×X (α) = SqP [Pr ] × SqX [X] .
r
It follows from Example 58.6 and (60.1) that

r¡ ¢ ¡ ¢
SqP [Pr ] = c(TPr )−1 [Pr ] = (1 + h)−r−1 ,
61. STEENROD OPERATIONS ON SMOOTH SCHEMES 289
where h = c1 (L) is the class of a hyperplane in Pr . By Proposition 50.4,

¡ ¢ ¡ ¢
p∗ SqP ×X (α) = deg(1 + h)−r−1 · SqX [X] .
r
We have µ ¶ µ ¶
−r−1 −r − 1 r 2r
deg(1 + h) = = (−1)
r r
and the latter binomial coefficient is even if r > 0. ¤
Theorem 60.5. Let f : Y → X be a projective morphism. Then the diagram
SqY
Ch(Y ) −−−→ Ch(Y )
 

f∗ y

f∗ y
SqX
Ch(X) −−−→ Ch(X)
is commutative.
Proof. The projective morphism f factors as the composition of a closed embedding
Y → Pr × X and the projection Pr × X → X, so the statement follows from Lemmas
60.3 and 60.4. ¤
Theorem 60.6. SqX X
k = 0 if k < 0 and Sq0 is the identity.
Proof. First suppose that X is a variety of dimension d. By dimension count, the

class SqX X
k ([X]) = Sgd−k (TX ) is trivial if k < 0. To compute Sq0 ([X]), we can extend the
base field to a perfect one and replace X by a smooth open subscheme. Then by (60.1),
¡ ¢ ¡ ¢
SqX 0 [X] = c0 (−TX ) [X] = [X],
i.e., SqX
0 is the identity on Chd (X).
In general, let Z ⊂ X be a closed subvariety and let j : Z → X be the¡ closed ¢
X
embedding.
¡ Z ¢ Then by Lemma 60.3 and the first part of the proof, the class Sqk [Z] =
j∗ Sqk ([Z]) is trivial for k < 0 and is equal to [Z] ∈ Ch(X) if k = 0. ¤
61. Steenrod operations on smooth schemes

In this section, we define Steenrod operations of cohomological type and prove their
basic properties.
Lemma 61.1. Let f : Y → X be a regular closed embedding of schemes of codimension
r and g : UY /G → UX /G the closed embedding induced by f . Then g is a regular closed
embedding of codimension 2r and the following diagram
X v
CH(X) −−−→ CH(UX /G)
 
F  F
f y yg
Y v
CH(Y ) −−−→ CH(UY /G)
is commutative.
Proof. The closed embedding UY → UX is regular of codimension 2r and the mor-

phism UX → UX /G is faithfully flat. Hence g is also a regular closed embedding by
Proposition 104.11. Let p : N → Y be the normal bundle of f . The Gysin homomor-
phism f F is the composition of the deformation homomorphism σf : CH(X) → CH(N )
and the inverse to the pullback isomorphism p∗f : CH(Y ) → CH(N ) (cf. §55.A).
The normal bundle Nh of the closed embedding h : UY → UX is the restriction of the
vector bundle N 2 × A1 on UY .
Z(X) −−−→ Z(X 2 × A1 )G −−−→ Z(UX )G Z(UX /G)
   
σf σ 2 σ  σg
y y f ×1 y h y
(61.2) Z(N ) −−−→ Z(N 2 × A1 )G −−−→ Z(Nh )G Z(Nh /G)
x x x x
   
   
Z(Y ) −−−→ Z(Y 2 × A1 )G −−−→ Z(UY )G Z(UY /G)
where the first homomorphism in every row takes a cycle α to α2 ×[A1 ], the other unmarked
maps are pull-back homomorphisms with respect to flat morphisms and the equalities
follow from Proposition 59.2. ¡P ¢
The deformation homomorphism is defined by σf ni [Zi ] = [Cki ], where ki : Y ∩
Zi → Zi is the restriction of f by Proposition 52.7, so the commutativity of the upper left
square follows from the equality of cycles [Cki × Ckj ] = [Cki ×kj ] (cf. Proposition 104.7).
The two other top squares are commutative by Proposition 51.5. The commutativity of
the left bottom square follows from Propositions 57.7 and 57.11. The two other squares
are commutative by Proposition 49.18.
The normal bundle Nh is an open subscheme of UN and of N 2 × A1 . Let j : Nh → UN
and l : Nh /G → UN /G be the open embeddings. The following diagram of the pull-back
homomorphisms
Z(N ) −−−→ Z(N 2 × A1 )G −−−→ Z(UN )G Z(UN /G)
° °  
° °  ∗
° ° j∗y yl
Z(N ) −−−→ Z(N 2 × A1 )G −−−→ Z(Nh )G Z(Nh /G)
is commutative by Proposition 49.18. It follows from Lemma 59.5 that the composition
in the top row factors through the rational equivalence, hence so does the composition in
the bottom row and then in the middle row of the diagram (61.2). Therefore the diagram
(61.2) yields a commutative diagram
vX
CH(X) / CH(UX /G)
OOO
OOOσg
σf OOO
² OO'
l∗
CH(N ) / CH(UN /G) / CH(Nh /G)
O oo7
∼ ooo
p∗f o
ooo ∗
vY
ooo pg
CH(Y ) / CH(UY /G)
The lemma follows from the commutativity of this diagram. ¤
Let f : Y → X be a closed embedding of smooth schemes with the normal bundle

N → Y . Consider the diagram
j
P(TY ⊕ 1) −−−→ P(TX ⊕ 1)
 

py qy

f
Y −−−→ X.
Lemma 61.3. We have c(N ) ◦ f ∗ ◦ sgTX = sgTY ◦j ∗ .
¡ ¢
Proof. By the Projective Bundle ¡ Theorem
¢ 57.13, the group CH P (T X ⊕ 1) is gen-
k ∗
erated by the elements β = e(LX ) q (α) with k ≥ 0 and α ∈ CH(X). We have
¡ ¢
(61.4) e(LX )• (β) = e(LX )• q ∗ (α) .
¡ ¢
Since j ∗ (LX ) = LY and j ∗ ◦q ∗ = p∗ ◦f ∗ , we have j ∗ β = e(LTY )k p∗ (f ∗ (α)) by Proposition
53.3(2) and therefore
¡ ¢ ¡ ¢
(61.5) e(LY )• j ∗ (β) = e(LY )• ◦ p∗ f ∗ (α) .
¡ ¢ ¡ ¢−1
By Proposition 104.16, c(N ) ◦ s f ∗ (TX ) = c(N ) ◦ c f ∗ (TX ) = c(TY )−1 = s(TY ). It
follows from (61.4), (61.5), Propositions 54.7 and 58.4(2) that
c(N ) ◦ f ∗ ◦ sgTX (β) = c(N ) ◦ f ∗ ◦ q ∗ ◦ e(LX )• (β)
¡ ¢
= c(N ) ◦ f ∗ ◦ q ∗ ◦ e(LX )• q ∗ (α)
= c(N ) ◦ f ∗ ◦ s(TX )(α)
¡ ¢
= c(N ) ◦ s(f ∗ TX ) f ∗ (α)
¡ ¢
= s(TY ) f ∗ (α)
¡ ¢
= p∗ ◦ e(LY )• ◦ p∗ f ∗ (α)
¡ ¢
= p∗ ◦ e(LY )• j ∗ (β)
= sgTY ◦j ∗ (β). ¤
Proposition 61.6. Let f : Y → X be a closed embedding of smooth schemes with the
normal bundle N . Then c(N ) ◦ f ∗ ◦ SqX = SqY ◦f ∗ .
Proof. By Proposition 105.8, the schemes BY /G and BX /G are smooth. Let
j : P(TY ⊕ 1) → P(TX ⊕ 1) and h : BY /G → BX /G
be the closed embeddings induced by f . Let α ∈ Ch(X). Choose β ∈ Ch(BX /G)
2
satisfying β|(UX /G) = αG (cf. (59.4)). It follows from Proposition 55.19 and Lemma 61.1
that
¡ ∗ ¢ ¡ ¢2
h (β) |(UY /G) = f ∗ (α) G .
By Proposition 55.19 and Lemma 61.3,

c(N ) ◦ f ∗ ◦ SqX (α) = c(N ) ◦ f ∗ ◦ sgTX ◦iF
X (β)
= sgTY ◦j ∗ ◦ iF
X (β)
= sgTY ◦iF ∗
Y ◦ h (β)
= SqY ◦f ∗ (α). ¤
Let X be a smooth scheme. We define the Steenrod operations of cohomological type
by the formula
SqX = c(TX ) ◦ SqX .
We write SqkX for kth homogeneous part of SqX . Thus SqkX is an operation
SqkX : Ch∗ (X) → Ch∗+k (X).
Proposition 61.7¡ (Wu
¢ Formula). ¡Let ¢Z be a smooth closed subscheme of a smooth
scheme X. Then SqX [Z] = j∗ ◦ c(N ) [Z] , where N is the normal bundle of the closed
embedding j : Z → X.
Proof. By Proposition 54.5 and (60.1),

¡ ¢ ¡ ¢
SqX [Z] = c(TX ) ◦ SqX [Z]
¡ ¢
= c(TX ) ◦ j∗ ◦ c(−TZ ) [Z]
¡ ¢ ¡ ¢
= j∗ ◦ c i∗ (TX ) ◦ c(−TZ ) [Z]
¡ ¢
= j∗ ◦ c(N ) [Z]
¡ ¢
since c(TZ ) ◦ c(N ) = c j ∗ (TX ) . ¤
Theorem 61.8. Let f : Y → X be a morphism of smooth schemes. Then the diagram
Sq
Ch(X) −−−X→ Ch(X)
 

f ∗y
f ∗
y
Sq
Ch(Y ) −−−Y→ Ch(Y )
is commutative.
Proof. Suppose first that f is a closed embedding with normal bundle N . It follows
from Propositions 54.5(2) and 61.6 that
f ∗ ◦ SqX = f ∗ ◦ c(TX ) ◦ SqX
¡ ¢
= c f ∗ (TX ) ◦ f ∗ ◦ SqX
= c(TY ) ◦ c(N ) ◦ f ∗ ◦ SqX
= c(TY ) ◦ SqY ◦f ∗
= SqY ◦f ∗ .
Secondly, consider the case of the projection g : Y × X → X. Let Z ⊂ X be a closed

subvariety. By (60.1), Propositions 58.12, 57.11, Corollary 104.8, and Theorem 60.2, we
have g ∗ ([Z]) = [Y × Z] = [Y ] × [Z] and
¡ ¢ ¡ ¢ ¡ ¢
SqY ×X g ∗ ([Z]) = c TY ×X ◦ SqY ×X [Y × Z]
¡ ¡ ¢¡ ¢
= c TY ) × c(TX ) SqY ([Y ]) × SqX ([Z])
¡ ¡ ¢ ¡ ¢
= c TY ) ◦ SqY ([Y ]) × c(TX ) ◦ SqX ([Z])
¡ ¢
= [Y ] × SqX [Z]
¡ ¢
= g ∗ SqX ([Z]) .
In the general case, write f = g ◦ h where h = (idX , f ) : Y → Y × X is the closed
embedding and g : Y × X → X is the projection. Then by the above,
f ∗ ◦ SqX = h∗ ◦ g ∗ ◦ SqX = h∗ ◦ SqY ×X ◦g ∗ = SqY ◦h∗ ◦ g ∗ = SqY ◦f ∗ . ¤
Proposition 61.9. Let f : Y → X be a smooth projective morphism of smooth
schemes. Then
SqX ◦f∗ = f∗ ◦ c(−Tf ) ◦ SqY ,
where Tf is the relative tangent bundle of f .
Proof. It follows from the exactness of the sequence
0 → Tf → TY → f ∗ (TX ) → 0
that c(TY ) = c(Tf ) ◦ c(f ∗ TX ). By Proposition 54.5(1) and Theorem 60.5,
SqX ◦f∗ = c(TX ) ◦ SqX ◦f∗

= c(TX ) ◦ f∗ ◦ SqY
= c(TX ) ◦ f∗ ◦ c(−TY ) ◦ SqY
= f∗ ◦ c(f ∗ TX ) ◦ c(−TY ) ◦ SqY
= f∗ ◦ c(−Tf ) ◦ SqY . ¤
Let X be a smooth variety of dimension d and let Z ⊂ X be a closed subvariety.

Consider the closed embedding j : P(TZ ⊕ 1) → P(TX ⊕ 1). By the Projective Bundle
Theorem 57.13 applied to the vector bundle TX ⊕ 1 over X of rank d + 1, there are unique
elements α0 , α1 , . . . , αd ∈ Ch(X) such that
d
¡ ¢ X ¡ ¢
j∗ [P(TZ ⊕ 1)] = e(L)k q ∗ (αk )
k=0
¡ ¢
in Ch P(TX ⊕ 1) , where L is the canonical line bundle over P(TX ⊕ 1) and q : P(TX ⊕ 1) →
X is the natural morphism. We set α := α0 + α1 + · · · + αd ∈ Ch(X).
¡ ¢
Lemma 61.10. SqX [Z] = s(TX )(α).
Proof. Let p : P(TZ ⊕ 1) → Z be the projection and i : Z → X the closed embedding,

so that i ◦ p = q ◦ j. The canonical line bundle L0 over P(TZ ⊕ 1) coincides with j ∗ (L)
and by Proposition 53.3,
¡ ¢
SqX [Z] = i∗ Sg(TZ )
¡ ¢
= i∗ ◦ p∗ ◦ e(L0 )• [P(TZ ⊕ 1)]
¡ ¢• ¡ ¢
= q∗ ◦ j∗ ◦ e j ∗ (L) [P(TZ ⊕ 1)]
¢
= q∗ ◦ e(L)• ◦ j∗ ([P(TZ ⊕ 1)]
d
X
•
¡ ¢
= q∗ ◦ e(L) ◦ e(L)k q ∗ (αk )
k=0
• ∗
= q∗ ◦ e(L) ◦ q (α)
= s(TX )(α). ¤
¡ ¢
Corollary 61.11. SqX [Z] = α in Ch(X).
Proof. By Lemma 61.10 and Proposition 58.9,
¡ ¢ ¡ ¢
SqX [Z] = c(TX ) SqX ([Z]) = c(TX )s(TX )(α) = α. ¤
Theorem 61.12. Let X be a smooth scheme. Then for any β ∈ Chk (X),


β if r = 0
r
SqX (β) = β 2 if r = k

0 if r < 0 or r > k.
Proof. By definition and Theorem 60.6, we have SqkX = 0 if k < 0 and Sq0X is the
identity operation.
We may assume that X is a variety and β = [Z] where Z ⊂ X is a closed subvariety
of codimension k. Since αi ∈ Ch2k−i (X), we have SqrX (β) = αk−r by Corollary 61.11.
Therefore, SqrX (β) = 0 if r > k.
As SqkX (β) = α0 , it remains to prove that β 2 = α0 . Consider the diagonal embedding
d : X → X 2 and the closed embedding h : TZ → TX . By the definition of the product in
Ch(X) and Proposition 52.7,
¡ ¢ ¡ ¢ ¡ ¢
p∗ (β 2 ) = p∗ ◦ d∗X [Z 2 ] = σd [Z 2 ] = h∗ [TZ ] ∈ Ch(TX ),
where p : TX → X is the canonical morphism. Let j : TX → P(TX ⊕ 1) be the open
embedding. Since the pullback j ∗ (L) of the canonical line bundle L over P(TX ⊕ 1) is a
trivial line bundle over TX , we have
(
¡ ¢ ¡ ¢ s ¡ ¢ p∗ (α) if s = 0,
j ∗ ◦ e(L)s q ∗ (α) = e j ∗ (L) j ∗ ◦ q ∗ (α) =
0 if s > 0
for every α ∈ Ch(X). Hence
£ ¤ £ ¤
p∗ (β 2 ) = TZ = j ∗ ( P(TZ ⊕ 1) ) = p∗ (α0 ),
therefore, β 2 = α0 since p∗ is an isomorphism. ¤
Theorem 61.13. Let X and Y be two smooth schemes. Then SqX×Y = SqX × SqY .
Proof. By Corollary 104.8, we have TX×Y = TX × TY . It follows from Theorem 60.2
and Proposition 58.12 that
SqX×Y = c(TX×Y ) ◦ SqX×Y
¡ ¢ ¡ ¢
= c(TX ) ◦ SqX × c(TY ) ◦ SqY
= SqX × SqY . ¤
Corollary 61.14 (Cartan Formula). Let X be a smooth scheme. Then SqX (α · β) =
SqX (α) · SqX (β) for all α, β ∈ Ch(X). Equivalently,
X
SqnX (α · β) = SqkX (α) · Sqm
X (β)
k+m=n
for all n.
Proof. Let i : X → X × X be the diagonal embedding. Then by Theorems 61.8 and

61.13,
¡ ¢
SqX (α · β) = SqX i∗ (α × β)
= i∗ SqX×Y (α × β)
¡ ¢
= i∗ SqX (α) × SqX (β)
= SqX (α) · SqX (β). ¤
Example 61.15. Let X = Pd be projective space and h ∈ Ch1 (X) the class of a
hyperplane. By Theorem 61.12, we have SqX (h) = h + h2 = h(1 + h). It follows from
Corollary 61.14 that
µ ¶
i i i r i i i+r
SqX (h ) = h (1 + h) , SqX (h ) = h .
r
By Example 104.20, the class of the tangent bundle TX is equal to (d + 1)[L] − 1, where
L is the canonical line bundle over X. Hence
c(TX ) = c(L)d+1 = (1 + h)d+1
and
SqX (hi ) = c(TX )−1 ◦ SqX (hi ) = hi (1 + h)i−d−1 .
CHAPTER XII
Category of Chow motives
In this chapter we study Chow motives. The notion of a Chow motive is due to
Grothendieck. Many (co)homology theories defined on the category Sm(F ) of smooth
complete schemes, such as Chow groups and more generally K-(co)homology groups take
values in the category of abelian groups. But the category Sm(F ) itself does not have the
structure of an additive category as we cannot add morphisms of schemes.
In this chapter, for an arbitrary commutative ring Λ, we shall construct the additive
categories of correspondences CR(F, Λ), CR∗ (F, Λ) and motives CM(F, Λ), CM∗ (F, Λ)
together with functors
Sm(F ) −−−→ CR(F, Λ) −−−→ CM(F, Λ)
 
 
y y
CR∗ (F, Λ) −−−→ CM∗ (F, Λ)
so that the theories with values in the category of abelian groups mentioned above fac-
tor through them. All of these the new categories do have the additional structure of
additive category. This makes them easier to work with than with the category Sm(F ).
Applications of these categories can be found in Chapter XVII later in this book.
Some classical theorems also have motivic analogs. For example, the Projective Bundle
Theorem 53.10 has such an analog (cf. Theorem 63.10) below. We shall se that the motive
of a projective bundle splits into a direct sum of certain motives already in the category
of correspondences CR(F, Λ). From this the classical Projective Bundle Theorem 53.10
can be obtained by applying an appropriate functor to the decomposition in CR(F, Λ).
In this chapter scheme means a separated scheme of finite type over a field.
62. Correspondences
A correspondence between two schemes X and Y is an element of CH(X × Y ). For
example, the graph of a morphism between X and Y is a correspondence. In this section
we study functorial properties of correspondences.
For a scheme Y over F , we have two canonical morphisms: the structure morphism
pY : Y → Spec(F ) and the diagonal closed embedding dY : Y → Y × Y . If Y is complete,
the map pY is proper and if Y is smooth, the closed embedding dY is regular.
Let X, Y and Z be schemes over F with Y proper and smooth. We consider morphisms
XZ
pY := 1X × pY × 1Z : X × Y × Z → X × Z
and
X Z
dY := 1X × dY × 1Z : X × Y × Z → X × Y × Y × Z.
If X = Spec(F ), we shall simply write pZY and dZY .
297
298 XII. CATEGORY OF CHOW MOTIVES
Define a bilinear pairing of K-homology groups (cf. §52)

A∗ (Y × Z, K∗ ) × A∗ (X × Y, K∗ ) → A∗ (X × Z, K∗ )
by
(62.1) (β, α) 7→ β ◦ α := (XpZY )∗ ◦ (XdZY )F (α × β).
For an element α ∈ A∗ (X × Y, K∗ ) we write αt for its image in A∗ (Y × X, K∗ ) under
the exchange isomorphism X × Y ' Y × X. The element αt is called the transpose of α.
By the definition of the pairing,
(β ◦ α)t = αt ◦ β t .
Proposition 62.2. The pairing (62.1) is associative. More precisely, for any four
schemes X, Y, Z, T over F with Y and Z smooth and complete and any
α ∈ A∗ (X × Y, K∗ ), β ∈ A∗ (Y × Z, K∗ ), and γ ∈ A∗ (Z × T, K∗ ), we have
(γ ◦ β) ◦ α = (XpTY ×Z )∗ ◦ (XdTY ×Z )F (α × β × γ) = γ ◦ (β ◦ α).
Proof. We prove the first equality. It follows from Corollary 55.4 that
(XdTY )F ◦ (X×Y ×YpTZ )∗ = (X×YpTZ )∗ ◦ (XdZ×T
Y )F .
By Propositions 50.4, 50.5, and 55.1, we have
¡ ¢
(γ ◦ β) ◦ α = (XpTY )∗ ◦ (XdTY )F α × (YpTZ )∗ (YdTZ )F (β × γ)
= (XpTY )∗ ◦ (XdTY )F ◦ (X×Y ×YpTZ )∗ ◦ (X×Y ×YdTZ )F (α × β × γ)
= (XpTY )∗ ◦ (X×YpTZ )∗ ◦ (XdZ×T
Y )F ◦ (X×Y ×YdTZ )F (α × β × γ)
= (XpTY ×Z )∗ ◦ (XdTY ×Z )F (α × β × γ). ¤
Let f : X → Y be a morphism of schemes. The isomorphic image of X under the
closed embedding (1X , f ) : X → X × Y is called the graph of f and is denoted by
Γf . Thus, Γf is a closed subscheme of X × Y isomorphic to X under the projection
X × Y → X. The class [Γf ] belongs to CH(X × Y ).
Proposition 62.3. Let X, Y, Z be schemes over F with Y smooth and complete.
(1) For every morphism g : Y → Z and α ∈ A∗ (X × Y, K∗ ),
[Γg ] ◦ α = (1X × g)∗ (α).
(2) For every morphism f : X → Y and β ∈ A∗ (Y × Z, K∗ ),
β ◦ [Γf ] = (f × 1Z )∗ (β).
Proof. (1): Consider the commutative diagram
X×Yp Xd
X × Y ←−−−Y− X ×Y ×Y Y
←−− − X ×Y
 

ry

ty
XdZ XpZ
Y
X × Y × Y × Z ←−− − X × Y × Z −−−Y→ X × Z
where r = 1X×Y × (1Y , g) and t = 1X × (1Y , g).
62. CORRESPONDENCES 299
The composition X×YpY ◦XdY is the identity on X × Y and XpZY ◦ t = 1X × g. It follows

from Corollary 55.4 that (XdZY )F ◦ r∗ = t∗ ◦ (XdY )F . We have
¡ ¢
[Γg ] ◦ α = (XpZY )∗ ◦ (XdZY )F α × [Γg ]
¡ ¢
= (XpZY )∗ ◦ (XdZY )F ◦ r∗ α × [Y ]
= (XpZY )∗ ◦ (XdZY )F ◦ r∗ ◦ (X×YpY )∗ (α)
= (XpZY )∗ ◦ t∗ ◦ (XdY )F ◦ (X×YpY )∗ (α)
= (1X × g)∗ (α).
(2): Consider the commutative diagram
pY ×Z v
Y × Z ←X−−− X ×Y ×Z ←−−− X ×Z
 

uy

vy
XdZ XpZ
Y
X × Y × Y × Z ←−− − X × Y × Z −−−Y→ X × Z
where u = (1X , f ) × 1Y ×Z and v = (1X , f ) × 1Z .
The composition XpZY ◦ v is the identity on X × Z and pYX×Z ◦ v = f × 1Z . It follows
from Corollary 55.4 that (XdZY )F ◦ u∗ = v∗ ◦ v ∗ . We have
¡ ¢
β ◦ [Γf ] = (XpZY )∗ ◦ (XdZY )F [Γf ] × β
¡ ¢
= (XpZY )∗ ◦ (XdZY )F ◦ u∗ [X] × β
= (XpZY )∗ ◦ (XdZY )F ◦ u∗ ◦ (pYX×Z )∗ (β)
= (XpZY )∗ ◦ v∗ ◦ v ∗ ◦ (pYX×Z )∗ (β)
= (f × 1Z )∗ (β). ¤
Corollary 62.4. Let X and Y be schemes over F and α ∈ A∗ (X × Y, K∗ ). Then
(1) If Y is smooth and complete then α ◦ [Γ1Y ] = α.
(2) If X is smooth and complete then [Γ1X ] ◦ α = α.
Corollary 62.5. Let f : X → Y and g : Y → Z be two morphisms. If Y is smooth
and complete then [Γg ] ◦ [Γf ] = [Γgf ].
Proof. By Proposition 62.3(1),
¡ ¢
[Γg ] ◦ [Γf ] = (1X × g)∗ [Γf ]
¡ ¢
= (1X × g)∗ (1X , f )∗ [X]
¡ ¢
= (1X , gf )∗ [X]
= [Γgf ]. ¤
Let X, Y and Z be arbitrary schemes and α ∈ A∗ (X × Y, K∗ ). If X is smooth and
complete, we have a well defined homomorphism
α∗ : A∗ (Z × X, K∗ ) → A∗ (Z × Y, K∗ ), β 7→ α ◦ β.
If α = [Γf ] with f : X → Y a morphism, it follows from Proposition 62.3(1) that

α∗ = (1Z × f )∗ .
If X is smooth and Z = Spec(F ), we get a homomorphism α∗ : A∗ (X, K∗ ) →
A∗ (Y, K∗ ). We have simpler formula for α∗ in the following case:
Proposition 62.6. Let α = [T ] with T ⊂ X × Y a closed subscheme with X smooth
and complete. Then α∗ = q∗ ◦ p∗ , where p : T → X and q : T → Y are the projections.
Proof. Let r : X × Y → Y be the projection, i : T → X × Y the closed embedding,
and f : T → X × T the graph of the projection p. Consider the commutative diagram
f q
X ×T ←−−− T −−−→ Y
  °

1X ×iy
 °
iy °
dY r
X × X × Y ←−X−− X × Y −−−→ Y.
It follows from Corollary 55.4 that i∗ ◦ f ∗ = (dYX )F ◦ (1X × i)∗ . Therefore for every
β ∈ A∗ (X, K∗ ), we have
α∗ (β) = r∗ ◦ (dYX )F (β × α)
¡ ¢
= r∗ ◦ (dYX )F ◦ (1X × i)∗ β × [T ]
¡ ¢
= r∗ ◦ i∗ ◦ f ∗ β × [T ]
¡ ¢
= q∗ ◦ f ∗ β × [T ]
= q∗ ◦ p∗ (β). ¤
If Y is smooth and complete, we have a well defined homomorphism
α∗ : A∗ (Y × Z, K∗ ) → A∗ (X × Z, K∗ ), β 7→ β ◦ α.
If α = [Γf ] for a flat morphism f : X → Y , it follows from Proposition 62.3(2) that
α∗ = (f × 1Z )∗ .
Let X, Y and Z be arbitrary schemes, α ∈ A∗ (X × Y, K∗ ), and g : Y → Z a proper
morphism. We define the composition of g and α by
g ◦ α := (1X × g)∗ (α) ∈ A∗ (X × Z, K∗ ).
If g ◦ α = [Γh ] for some morphism h : X → Z, we abuse notation and just write g ◦ α = h.
If Y is smooth and complete, we have g ◦ α = [Γg ] ◦ α by Proposition 62.3(1).
Similarly, if β ∈ A∗ (Y × Z, K∗ ) and f : X → Y is a flat morphism, we define the
composition of β and f by
β ◦ f := (f × 1Z )∗ (β) ∈ A∗ (X × Z, K∗ ).
If Y is smooth and complete, we have β ◦ f = β ◦ [Γf ] by Proposition 62.3(2).
The following statement is an analogue of Proposition 62.2 with less assumptions on
the schemes.
Proposition 62.7. Let X, Y , Z and T be arbitrary schemes.
63. CATEGORIES OF CORRESPONDENCES 301
(1) Let α ∈ A∗ (X × Y, K∗ ), γ ∈ A∗ (T × X, K∗ ), and g : Y → Z a proper morphism.

If X is smooth and complete then
(g ◦ α) ◦ γ = g ◦ (α ◦ γ), i.e., (g ◦ α)∗ = g∗ ◦ α∗ .
(2) Let β ∈ A∗ (Y × Z, K∗ ), δ ∈ A∗ (Z × T, K∗ ), and f : X → Y a flat morphism. If
Z is smooth and complete then
δ ◦ (β ◦ f ) = (δ ◦ β) ◦ f, i.e., (β ◦ f )∗ = f ∗ ◦ β ∗ .
Proof. (1): Consider the commutative diagram with fiber product squares
TdY TpY
X X
T × X × X × Y ←−−− T × X × Y −−−→ T ×Y
  

1T ×X×X ×g y

1T ×X ×g y
1 ×g
yT
TdZ TpZ
X X
T × X × X × Z ←−− − T × X × Z −−− → T × Z.
It follows from Proposition 50.4 and Corollary 55.4 that
g ◦ (α ◦ γ) = (1T × g)∗ (α ◦ γ)
= (1T × g)∗ ◦ (TpYX )∗ ◦ (TdYX )F (γ × α)
= (TpZX )∗ ◦ (TdZX )F ◦ (1T ×X×X × g)∗ (γ × α)
= (1X × g)∗ (α) ◦ γ
= (g ◦ α) ◦ γ.
(2): The proof is similar using Propositions 49.20, 50.5 and 55.5 instead. ¤
If γ ∈ A∗ (Y × X, K∗ ) and g : Y → Z is a proper morphism, we write γ ◦ g t for
(g ◦ γ t )t ∈ A∗ (Z × X, K∗ ). Similarly, if δ ∈ A∗ (Z × Y, K∗ ) and f : X → Y is a flat
morphism, we define the composition f t ◦ δ ∈ A∗ (Z × X, K∗ ) to be (δ t ◦ f )t .
63. Categories of correspondences

Let Λ be a commutative ring. For a scheme Z, we write CH(Z; Λ) for the Λ-module
CH(Z) ⊗ Λ.
Let X and Y be smooth complete schemes over F . Let X1 , X2 , . . . , Xn be the irre-
ducible components of X of dimension d1 , d2 , . . . , dn respectively. For every i ∈ Z, we set
n
a
Corri (X, Y ; Λ) := CHi+dk (Xk × Y ; Λ).
k=1
An element α ∈ Corri (X, Y ; Λ) is called a correspondence between X and Y of degree i
with coefficients in Λ. We write α : X Ã Y .
Let Z be another smooth complete scheme. By Proposition 62.2, the bilinear pairing
(β, α) 7→ β ◦ α on Chow groups yields an associative pairing (composition)
(63.1) Corri (Y, Z; Λ) × Corrj (X, Y ; Λ) → Corri+j (X, Z; Λ).
The following proposition gives an alternative formula for this composition involving
only projection morphisms.
¡ ¢
Proposition 63.2. β ◦ α = (XpZY )∗ (X×YpZ )∗ (α) · (pYX×Z )∗ (β) .
Proof. Let f : X ×Y ×Y ×Z → X ×Y ×Z ×X ×Y ×Z be defined by f (x, y, y 0 , z) =
(x, y, z, x, y 0 , z). We have f ◦XdZY = dX×Y ×Z , therefore
β ◦ α = (XpZY )∗ ◦ (XdZY )F (α × β)
¡ ¢
= (XpZY )∗ ◦ (XdZY )F ◦ f ∗ α × [Z] × [X] × β
¡ ¢
= (XpZY )∗ ◦ (dX×Y ×Z )F (X×YpZ )∗ (α) × (pYX×Z )∗ (β)
¡ ¢
= (XpZY )∗ (X×YpZ )∗ (α) · (pYX×Z )∗ (β) . ¤
Define the category CR∗ (F, Λ) of correspondences with coefficients in Λ over F as
follows: Objects of CR∗ (F, Λ) are smooth complete schemes over F . A morphism between
X and Y is an element of the graded group
a
Corrk (X, Y ; Λ).
k∈Z
Composition of morphisms is given by (63.1). The identity morphism of X in CR∗ (F, Λ) is

Γid ⊗1, where Γid is the class of the graph of the identity morphism 1X (cf. Corollary 62.4).
The direct sum in CR∗ (F, Λ) is given by the disjoint union of schemes. As the composition
law in CR∗ (F, Λ) is bilinear and associative by Proposition 62.2, the category CR∗ (F, Λ)
is additive. Abusing notation, we write Λ for the object Spec(F ) in this category.
An object of CR∗ (F, Λ) is called a Chow motive or simply a motive. If X is a smooth
complete scheme we write M (X) for it as an object in CR∗ (F, Λ).
We define another category C(F, Λ) as follows. Objects of C(F, Λ) are pairs (X, i),
where X is a smooth complete scheme over F and i ∈ Z. A morphism between (X, i)
and (Y, j) is an element of Corri−j (X, Y ; Λ). The composition of morphisms is given by
(63.1). The morphisms between two objects form an abelian group and the composition is
bilinear and associative by Proposition 62.2, therefore, C(F, Λ) is a preadditive category.
There is an additive functor C(F, Λ) → CR∗ (F, Λ) taking an object (X, i) to X and
that is the natural inclusion on morphisms.
Let A be a preadditive category. Define the additive completion of A to be the
following ` e its objects are finite sequences of objects A1 , . . . , An of A written in
category A: ` `
the form i=1 Ai . A morphism between ni=1 Ai and m
n
j=1 Bj is given by an n × m-matrix
of morphisms Ai → Bj . The composition of morphisms is the matrix multiplication. The
category Ae has finite products and coproducts and is therefore an additive category. The
category A is a full subcategory of A.e
Let CR(F, Λ) denote the additive completion of C(F, Λ). We call it the category of
graded correspondences with coefficients in Λ over F . An object of CR(F, Λ) is also called
a Chow motive or simply a motive. We will write M (X)(i) for (X, i) and simply M (X)
for (X, i)for (X, 0). The functor C(F, Λ) → CR∗ (F, Λ) extends naturally to an additive
functor
(63.3) CR(F, Λ) → CR∗ (F, Λ).
taking M (X)(i) to M (X). Write Λ(i) for the motive (Spec F, i) in CR(F, Λ). The motives
Λ(i) in CR(F, Λ) and Λ in CR∗ (F, Λ) are called the Tate motives.
63. CATEGORIES OF CORRESPONDENCES 303
Functor (63.3) is faithful but not full. Nevertheless it has the following nice property.
Proposition 63.4. Let f be a morphism in CR(F, Λ). If the image of f in CR∗ (F, Λ)
is an isomorphism then f itself is an isomorphism.
` `
Proof. Let f be a morphism between the objects ni=1 Xi (ai ) and m j=1 Yj (bj ). Thus
f is given by an n × m matrix A = (fij ) with fij ∈ Corraj −bi (Xj , Yi ; Λ). Let B = (gkl )
be the matrix of the inverse of f in CR∗ (F, Λ), so that gkl ∈ Corr∗ (Yl , Xk ; Λ). Let ḡkl be
the homogeneous component of gkl of degree bl − ak and B = (ḡkl ). As AB = AB and
BA = BA are the identity matrices, we have B = B = A−1 . Therefore, B is the matrix
of the inverse of f in CR(F, Λ). ¤
A ring homomorphism Λ → Λ0 gives rise to natural functors CR∗ (F, Λ) → CR∗ (F, Λ0 )
and CR(F, Λ) → CR(F, Λ0 ) that are identical on objects. We simply write CR∗ (F ) for
CR∗ (F, Z) and CR(F ) for CR(F, Z).
It follows from Corollary 62.5 that there is a functor
Sm(F ) → CR(F, Λ)
taking a smooth complete scheme
¡ X to M (X)
¢ and a morphism f : X → Y to [Γf ] ⊗ 1 in
Corr0 (X, Y ; Λ) = MorCR(F,Λ) M (X), M (Y ) , where Γf is the graph of f .
Let X and Y be smooth complete schemes and i, j ∈ Z. We have
¡ ¢
HomCR(F ) M (X)(i), M (Y )(j) = Corri−j (X, Y ; Λ).
In particular,
¡ ¢
(63.5) HomCR(F,Λ) Λ(i), M (X) = CHi (X; Λ),
¡ ¢
(63.6) HomCR(F,Λ) M (X), Λ(i) = CHi (X; Λ).
The category CR(F, Λ) has the structure of a tensor category given by
M (X)(i) ⊗ M (Y )(j) := M (X × Y )(i + j).
In particular,
M (X)(i) ⊗ Λ(j) = M (X)(i + j).
Let Y be a smooth variety of dimension d. By the definition of a morphism in CR(F ),
the equality
¡ ¢ ¡ ¢
(63.7) HomCR(F,Λ) M (Y )(i), N = CHd+i M (Y ) ⊗ N ; Λ
holds for every N of the form M (X)(j), where X is a smooth complete scheme; and,
therefore, by additivity it holds for all motives N . Similarly,
¡ ¢ ¡ ¢
(63.8) HomCR(F,Λ) N, M (Y )(i) = CHd+i N ⊗ M (Y ); Λ .
The following statement is a variant of the Yoneda Lemma.
Lemma 63.9. Let α : N → P be a morphism in CR(F, Λ). Then the following
(1) α is an isomorphism.
(2) For every smooth complete scheme Y , the homomorphism

¡ ¢ ¡ ¢
(1Y ⊗ α)∗ : CH∗ M (Y ) ⊗ N ; Λ → CH∗ M (Y ) ⊗ P ; Λ
is an isomorphism.
(3) For every smooth complete scheme X, the homomorphism
¡ ¢ ¡ ¢
(1Y ⊗ α)∗ : CH∗ M (Y ) ⊗ P ; Λ → CH∗ M (Y ) ⊗ N ; Λ
is an isomorphism.
Proof. Clearly (1) ⇒ (2) and (1) ⇒ (3). We prove that (2) implies (1) (the proof
of the implication (3) ⇒ (1) is similar and left to the reader). It follows from (63.7) that
the natural homomorphism
HomCR(F,Λ) (M, N ) → HomCR(F,Λ) (M, P )
is an isomorphism if M = M (Y )(i) for any smooth complete variety Y . By additivity, it
is isomorphism for all motives M . The statement now follows from the (usual) Yoneda
Lemma. ¤
The following statement is the motivic version of the Projective Bundle Theorem
53.10.
Theorem 63.10. Let E → X¡ be a¢vector`bundle of rank r over a smooth complete
scheme X. Then the motives M P(E) and r−1 i=0 M (X)(i) are naturally isomorphic in
CR(F, Λ).
Proof. Let Y be a smooth complete scheme over F . Applying the Projective Bundle
Theorem
`r−1 53.10 to the vector
¡ bundle¢ E × Y → X × Y , we see that the Chow groups of
i=0 M (X ×Y )(i) and M P(E)×Y are isomorphic. Moreover, in view of Remark 53.11,
this isomorphism is natural in Y with respect to morphisms in the category
`r−1 CR(F, Λ).
In ¡other ¢words, the functors on CR(F, Λ) represented by the objects i=0 M (X)(i) and
M P(E) are isomorphic. By the Yoneda Lemma, the objects are isomorphic in CR(F, Λ).
¤
¡ ¢
Corollary 63.11. The motive M P(E) is isomorphic to the direct sum M (X)r of
r copies of M (X) in the category CR∗ (F, Λ).
64. Category of Chow motives

Let A be an additive category. An idempotent e : A → A in A is called split if
∼
there is an isomorphism f : A → B ⊕ C such that e coincides with the composition
f p i f −1
A− →B⊕C − →B− → B ⊕ C −−→ A, where p and i are the canonical morphisms.
The idempotent completion of an additive category A is the category A defined as
follows. Objects of A are the pairs (A, e), where A is an object of A and e : A → A is an
idempotent. The group of morphisms between (A, e) and (B, f ) is f ◦ HomA (A, B) ◦ e.
Every idempotent in A is split.
The assignment A 7→ (A, 1A ) defines a full and faithful functor from A to A. We
identify A with a full subcategory of A.
65. DUALITY 305
Let Λ be a commutative ring. The idempotent completion of the category CR(F, Λ)

is called the category of graded Chow motives with coefficients in Λ and is denoted by
CM(F, Λ). By definition, every object of CM(F, Λ) is a direct summand of a finite direct
sum of motives of the form M (X)(i), where X is a smooth complete scheme over F . We
write CM(F ) for CM(F, Z).
Similarly, the idempotent completion CM∗ (F, Λ) of CR∗ (F, Λ) is called the category
of Chow motives with coefficients in Λ. Note that Proposition 63.4 holds for the natural
functor CM(F, Λ) → CM∗ (F, Λ).
We have functors
Sm(F ) → CR(F, Λ) → CM(F, Λ).
The second functor is full and faithful, so we can view CR(F, Λ) as a full subcategory of
CM(F, Λ) which we do. Note that CM(F, Λ) inherits the structure of a tensor category.
An object of CM(F, Λ) is also called a motive. We will keep the same notation
M (X)(i), Λ(i), etc. for the corresponding motives in CM(F, Λ). The motives Λ(i) and Λ
are called the Tate motives.
We use formulas (63.5) and (63.6) in order to define Chow groups with coefficients in
Λ for an arbitrary motive M :
¡ ¢ ¡ ¢
CHi (M ; Λ) := HomCM(F,Λ) Λ(i), M , CHi (M ; Λ) := HomCM(F,Λ) M, Λ(i) .
The functor from CM(F, Λ) to the category of Λ-modules, taking a motive M to CHi (M ; Λ)
(respectively the cofunctor M 7→ CHi (M ; Λ)) is then represented (respectively corepre-
sented) by Λ(i).
Let Y be a smooth variety of dimension d. It follows from 63.7 and 63.8 that
¡ ¢ ¡ ¢
(64.1) HomCM(F,Λ) M (Y )(i), N = CHd+i M (Y ) ⊗ N ; Λ ,
¡ ¢ ¡ ¢
HomCM(F,Λ) N, M (Y )(i) = CHd+i N ⊗ M (Y ); Λ .
for all motives N in CM(F, Λ).
Let M and N be objects in CM(F ). The tensor product of two morphisms M → Λ(i)
and N → Λ(j) defines a pairing
(64.2) CH∗ (M ; Λ) ⊗ CH∗ (N ; Λ) → CH∗ (M ⊗ N ; Λ).
Note that this is an isomorphism whenever M (or N ) is a Tate motive.
We say that an object M of CR(F, Λ) is split if M is isomorphic to a (finite) coproduct
of Tate motives. The additivity property of the pairing yields
Proposition 64.3. Let M (or N ) be a split motive. Then the homomorphism (64.2)
is an isomorphism.
65. Duality
There is an additive duality functor ∗ : CM(F, Λ)op → CM(F, Λ) uniquely determined
by the rule M (X)(i)∗ = M (X)(−d−i) where X is a smooth complete variety of dimension
d and α∗ = αt for any correspondence α. In particular, Λ(i)∗ = Λ(−i). The composition
∗ ◦ ∗ is the identity functor.
It follows from the definition of the duality functor that
HomCM(F,Λ) (M ∗ , N ∗ ) = HomCM(F,Λ) (N, M )
for every two motives M and N . In particular, setting N = Λ(i), we get

CHi (M ∗ ; Λ) = CH−i (M ; Λ).
The equality (64.1) reads as follows:
¡ ¢ ¡ ¢
(65.1) HomCM(F,Λ) M (Y )(i), N = CH0 M (Y )(i)∗ ⊗ N ; Λ
for every smooth complete scheme Y . Set
Hom(M, N ) := M ∗ ⊗ N
for every two motives M and N . By additivity, the equality (65.1) yields
¡ ¢
HomCM(F,Λ) (M, N ) = CH0 Hom(M, N ); Λ .
Since the duality functor commutes with the tensor product, the definition of Hom satisfies
the associativity law
¡ ¢
Hom(M ⊗ N, P ) = Hom M, Hom(N, P )
for all motives M , N and P . Applying CH0 we get
¡ ¢
HomCM(F,Λ) (M ⊗ N, P ) = HomCM(F,Λ) M, Hom(N, P ) .
66. Motives of cellular schemes

Motives of cellular schemes were considered in [71]. Recall that a morphism p : U → Y
over F is an affine bundle of rank d if f is flat and the fiber of p over any point y ∈ Y is
isomorphism to the affine space AdF (y) .
A scheme X over F is called (relatively) cellular if there is given a filtration by closed
subschemes
(66.1) ∅ = X−1 ⊂ X0 ⊂ X1 ⊂ · · · ⊂ Xn = X
together with affine bundles pi : Ui := Xi \ Xi−1 → Yi of rank di , for all i = 0, . . . , n,
where Yi is a smooth complete scheme, for all i ∈ [1, n] (we don’t assume that Y0 is either
smooth or complete).
The graph Γpi of the morphism pi is a subscheme of Ui × Yi . Let αi in CH(Xi × Yi )
be the class of the closure of Γpi in Xi × Yi . We have α0 = Γp0 .
We view αi as a correspondence Xi Ã Yi of degree 0. Let fi : Xi → X be the closed
embedding. The correspondence fi ◦ αit ∈ CH(Yi × X) between Yi and X is well defined
and of degree di for all i ≥ 1.
Let Z be a scheme over F . If h is a morphism of schemes, we still write h for the
morphism 1Z × h. We define homomorphisms
ai : CH∗−di (Z × Yi ) → CH∗ (Z × X)
for all i = 0, . . . , n as follows: We set a0 = (f0 )∗ ◦ (p∗0 )−1 and ai = (fi ◦ αit )∗ if i ≥ 1. If Y0
is smooth and complete then a0 = (f0 ◦ α0t )∗ .
Theorem 66.2. Let X be a cellular scheme with filtration (66.1). Then for every
scheme Z over F , the homomorphism
Xn n
a
(66.3) ai : CH∗−di (Z × Yi ) → CH∗ (Z × X)
i=0 i=0
66. MOTIVES OF CELLULAR SCHEMES 307
is an isomorphism.
Proof. Let gi : Ui → Xi denote the open embedding. By the definition of αi , we have
αi ◦ gi = pi . It follows from Proposition 62.7(2) that for every scheme Z, the composition
α∗
i i g∗
A(Yi × Z, K∗ ) −→ A(Xi × Z, K∗ ) −
→ A(Ui × Z, K∗ )
coincides with the pull-back homomorphism p∗i for i ≥ 1. By Theorem 52.13, the map
p∗i is an isomorphism. Hence gi∗ is a split surjection. Therefore, in the localization exact
sequence (cf. §52.D)
g∗
i δ
Ak+1 (Xi × Z, K−k ) −
→ Ak+1 (Ui × Z, K−k ) −
→
i g∗
CHk (Xi−1 × Z) → CHk (Xi × Z) −
→ CHk (Ui × Z) → 0
the connecting homomorphism δ is trivial. Consequently, we have a short exact sequence
i s
0 → CH(Xi−1 × Z) → CH(Xi × Z) −
→ CH(Yi × Z) → 0
where si = p∗i −1 ◦ gi∗ and si is split by αi∗ : CH(Yi × Z) → CH(Xi × Z). In particular,
CH(Xi ×Z) is isomorphic to CH(Xi−1 ×Z)⊕CH(Yi ×Z). Iterating, we see that CH(X ×Z)
is isomorphic to the coproduct of the CH(Yi × Z) over all i ∈ [0, n]. The inclusion of
CH(Yi × Z) into CH(X × Z) coincides with the composition
i α∗ (fi )∗
CH(Yi × Z) −→ CH(Xi × Z) −−→ CH(X × Z),
where α0∗ is understood to be p∗0 . If i ≥ 1, we have ai = (fi )∗ ◦ (αit )∗ by Proposition
62.7(1). Under the identification of CH(Yi × Z) with CH(Z × Yi ), we have (αit )∗ = αi∗ ,
hence ai = (fi )∗ ◦ αi∗ . It follows that the homomorphism (66.3) is an isomorphism. ¤
Lemma 63.9 yields
Corollary 66.4. Let X be a smooth complete cellular scheme with filtration (66.1)
and with all Yi smooth complete. Then the morphism
an
M (Yi )(di ) → M (X),
i=0
defined by the sequence of correspondences fi ◦ αit is an isomorphism in the category of

correspondences CR(F ).
Example 66.5. Let X = Pn . Consider the filtration given by Xi = Pi , i ∈ [0, n]. We
have Ui = Ai and Yi = Spec(F ). By Corollary 66.4,
M (Pn ) = Z ⊕ Z(1) ⊕ · · · ⊕ Z(n).
Example 66.6. Let ϕ be a non-degenerate quadratic form on V and X the associated
quadric of say dimension d. Consider the following filtration on X ×X: X0 is the image of
the diagonal embedding of X into X × X, X1 consists of all pairs of orthogonal isotropic
lines (L1 , L2 ), and X2 = X × X. We also set Y0 = X (with the identity projection of X0
on Y0 ), Y2 = X, and Y1 is the flag variety F l of pairs (L, P ), where L and P are a totally
isotropic line and plane respectively satisfying L ⊂ P .
We claim that the morphism p1 : U1 → Y1 taking a pair (L1 , L2 ) to (L1 , L1 + L2 )

is an affine bundle. To do this we use the criterion of Lemma 52.12. Let R be a local
commutative F -algebra. An R-point of Y1 is a pair (L, P ), where L and P are totally
isotropic direct summands of the R-module VR := V ⊗F R of rank 1 and 2 respectively
with L ⊂ P . Let {e, f } be an R-basis of P such ¡ that L =¢ Re. Then the morphism
A1R → Spec(R) ×Y1 U1 taking an a to the point L, R(ae + f ) of the fiber product is an
isomorphism. It follows from Lemma 52.12 that p1 is an affine bundle.
We claim that the first projection p2 : U2 → Y2 is an affine bundle of rank d. We
again apply the criterion of Lemma 52.12. Let R be a local commutative F -algebra. An
R-point of Y2 is a totally isotropic direct summand L ⊂ VR of rank 1. Choose a basis
of VR so that ϕR is given by a polynomial t0 t1 + ψ(T 0 ), where ψ is a quadratic form in
the variables T 0 = (t2 , . . . , td+1 ) over R, and the orthogonal complement L⊥ is given by
0
t0 = 0. Then the fiber product Spec(R) ×Y2 U2 is given by the equation tt01 + ψ( Tt0 ) = 0
and therefore is isomorphic to AdR . It follows by Lemma 52.12 that p2 is an affine bundle.
By Corollary 66.4, we conclude that
M (X × X) ' M (X) ⊕ M (F l)(1) ⊕ M (X)(d).
Example 66.7. Assume that the quadric X in Example 66.6 is isotropic. The cellular
structure on X × X is a structure “over X” in the sense that X × X itself as well as the
bases Yi of the cells have morphisms to X with affine bundles of the cellular structure
morphisms over X. Making the base change of the cellular structure with respect to an
F -point Spec(F ) → X of the isotropic quadric X corresponding to an isotropic line L,
we get a cellular structure on X given by the filtration X00 ⊂ X10 ⊂ X20 = X, where
X00 = {L} and X10 consists of all isotropic lines orthogonal to L. We have Y00 = Spec(F ),
Y10 is the quadric given by the quadratic form on L⊥ /L induced by ϕ, and Y20 = Spec(F ).
The quadric Y10 is isomorphic to a projective quadric Y of dimension d − 2, given by a
quadratic form Witt equivalent to ϕ. By Corollary 66.4,
M (X) ' Z ⊕ M (Y )(1) ⊕ Z(d).
67. Nilpotence Theorem
Let Λ be a commutative ring and Y a smooth complete scheme over F . For every
scheme X and elements α ∈ CH(Y × Y ; Λ) and β ∈ CH(X × Y ; Λ), the compositions
αk = α ◦ · · · ◦ α in CH(Y × Y ; Λ) and αk ◦ β in CH(X × Y ; Λ) are defined.
Theorem 67.1 (Nilpotence Theorem). Let Y be a smooth complete scheme and X
a scheme
¡ of¢dimension d over F . Let α ∈ CH(Y × Y ; Λ) be an element satisfying α ◦
CH YF (x) ; Λ = 0 for every x ∈ X. Then
αd+1 ◦ CH(X × Y ; Λ) = 0.
Proof. Consider the filtration

0 = C−1 ⊂ C0 ⊂ · · · ⊂ Cd = CH(X × Y ; Λ),
where Ci is the Λ-submodule of CH(X ×Y ; Λ) generated by the images of the push-forward
homomorphisms
CH(W × Y ; Λ) → CH(X × Y ; Λ),
67. NILPOTENCE THEOREM 309
for all closed subvarieties W ⊂ X of dimension at most k. It suffices to prove that

α ◦ Ck ⊂ Ck−1 for all k ∈ [0, d].
Let W be a closed subvariety of X of dimension k. Denote by i : W → X the closed
embedding and by w¡the generic
¢ point of W . Pick any element β ∈ CH(W × Y ; Λ). We
shall prove that α ◦ i∗ (β) ∈ Ck−1 . Let βw be the pull-back of β under the canonical
morphism YF (w) → W × Y . By assumption, α ◦ βw = 0. By the Continuity Property
52.9, there is a nonempty open subscheme U (a neighborhood of w) in W satisfying
α ◦ (β|U ×Y ) = 0. It follows by Proposition 62.7(2) that
(α ◦ β)|U ×Y = α ◦ (β|U ×Y ) = 0.
The complement V of U in W is a closed subscheme of W of dimension less than k. It
follows from the exactness of the localization sequence (cf. §52.D)
CH(V × Y ; Λ) → CH(W × Y ; Λ) → CH(U × Y ; Λ) → 0
that α ◦ β belongs to the image of the first map in this sequence. Therefore, the push-
forward of the element α ◦ β in CH(X × Y ; Λ) lies in the image of the push-forward
homomorphism
CH(V × Y ; Λ) → CH(X × Y ; Λ).
Consequently, α ◦ (i∗ β) = α ◦ (β ◦ it ) = (α ◦ β) ◦ it = (i × 1Y )∗ (α ◦ β) ∈ Ck−1 . ¤
The Nilpotence Theorem 67.1 was originally proven by Rost using cycle modules
technique.
Part
Quadratic forms and algebraic cycles

CHAPTER XIII
Cycles on powers of quadrics
Throughout this chapter, F is a field (of an arbitrary characteristic) and with the
exception of §70 and §71, X is a smooth projective quadric over F of even dimension
D = 2d ≥ 0 or of odd dimension D = 2d + 1 ≥ 1 given by a non-degenerate quadratic
form ϕ on a vector space V over F (of dimension D + 2). For any integer r ≥ 1, we write
X r for the direct product X × · · · × X (over F ) of r copies of X.
In this chapter, we study algebraic cycles on X r for all r. We also obtain many results
for the special case r = 2. In order to make their statements and proofs more accessible,
we introduce certain diagrams of cycles on X 2 (cf. §73).
For a, b ∈ Z, we write [a, b] (respectively [a, b)) for the set of integers c satisfying
a ≤ c ≤ b (respectively a ≤ c < b).
68. Split quadrics

In this section the quadric X will be split, i.e., the Witt index i0 (X) has the maximal
value d + 1.
Let V be the underlying vector space of ϕ. Fix a maximal totally isotropic subspace
W ⊂ V . We write P(V ) for the projective space of V ; this is the projective space in which
the quadric X lies as a hypersurface. Note that the subspace P(W ) of P(V ) is contained
in X.
Proposition
¡ ¢ 68.1. Let h ∈ CH1 (X) be the pull-back of the hyperplane class in
CH P(V ) . For any integer i ∈ [0, d], let li ∈ CHi (X) be the class of an i-dimensional
1
subspace of P(W ). Then the total Chow group CH(X) is free with basis {hi , li | i ∈ [0, d]}.
Moreover, the following multiplication rule holds in the ring CH(X):
h · li = li−1 for any i ∈ [1, d].
Proof. Let W ⊥ be the orthogonal complement of W in V (clearly, W ⊥ = W if D
is even; otherwise, W ⊥ contains W as a hyperplane). The quotient map V → V /W ⊥
induces a morphism X \ P(W ) → P(V /W ⊥ ), which is an affine bundle of rank D − d.
Therefore, by Theorem 66.2,
¡ ¢ ¡ ¢
CHi (X) ' CHi P(W ) ⊕ CHi−D+d P(V /W ⊥ )
¡ ¢
for any i, where the injection CH∗ P(W ) ,→ CH∗ (X) is the push-forward with respect
to the embedding P(W ) ,→ X.
To better understand the second summand in the decomposition of CH(X), we note
that the reduced intersection of P(W ⊥ ) with X in P(V ) is P(W ), and that the affine bundle
X \ P(W ) → P(V /W ⊥ ) above is the composite of the closed embedding X \ P(W ) ,→
P(V ) \ P(W ⊥ ) with the evident vector bundle P(V ) \ P(W ⊥ ) → P(V /W ⊥ ). It follows that
313
314 XIII. CYCLES ON POWERS OF QUADRICS
¡ ¢
for any i ≤ d ¡the image
¢ of CH i
P (V /W ⊥
) in CHi (X) coincides with the image of the
pull-back CHi P(V ) → CHi (X) (which is generated by hi ).
To check the multiplication formula, we consider ¡ ¢ closed embeddings f : P(W ) ,→ X
the
and g : X ,→ P(V ). Write Li for the class in CH ¡P(W ) ¢ of an i-dimensional linear subspace
of P(W ), and H for the hyperplane class in CH P(V ) . Since h = g ∗ (H) and li = f∗ (Li ),
we have by the projection formula (Proposition 56.9) and functoriality of the pull-back
(Proposition 55.18),
¡ ¢
h · li = g ∗ (H) · f∗ (Li ) = f∗ (f ◦ g)∗ (H) · Li .
By Corollary 57.19 (together ¡with Propositions
¢ 104.16 and 55.19), we see that (f ◦ g)∗ (H)
is the hyperplane class in CH P(W ) hence (f ◦ g)∗ (H) · Li = Li−1 by Example 57.22. ¤
Proposition 68.2. For each i ∈ [0, D/2), the i-dimensional subspaces of P(V ) lying
inside of X have the same class in CHi (X). If D is even there are precisely two different
classes of d-dimensional subspaces and the sum of these two classes is equal to hd .
¡ ¢
Proof. By Proposition 68.1, the push-forward homomorphism CHi (X) → CHi P(V )
is injective (even bijective) if i ¡∈ [0, D/2).
¢ Since the i-dimensional linear subspaces of
P(V ) have the same class in CH P(V ) , the first statement of Proposition 68.2 follows.
Assume that D is even. Then {hd , ld } is a basis for the group CHd (X), where ld is
the class of the special linear subspace P(W ) ⊂ X. Let ld0 ∈ CHd (X) be the class of
0
an arbitrary d-dimensional linear subspace of X. Since¡ ld and ¢ ld have the same image
under the push-forward homomorphism CHd (X) → CHd P(V ) whose kernel is generated
by hd − 2ld , one has ld0 = ld + n(hd − 2ld ) for some n ∈ Z. Since there exists a linear
automorphism of X moving ld to ld0 , and hd is of course invariant with respect to any linear
automorphism, hd and ld0 also form a basis for CHd (X); consequently, the determinant of
the matrix µ ¶
1 n
0 1 − 2n
is ±1, i.e., n is 0 or 1 and ld0 is ld or hd − ld . So there are at most two different rational
equivalence classes of d-dimensional linear subspaces of X and the sum of two different
classes (if they exist) is equal to hd .
Now let U be a d-codimensional subspace of V containing W (as a hyperplane). The
orthogonal complement U ⊥ has codimension 1 in W ⊥ = W , therefore codimU U ⊥ = 2.
The induced 2-dimensional quadratic form on U/U ⊥ is a hyperbolic plane. The corre-
sponding quadric consists of two points W/U ⊥ and W 0 /U ⊥ for a uniquely determined
maximal totally isotropic subspace W 0 ⊂ V . Moreover, the intersection X ∩ P(U ) is re-
duced and its irreducible components are P(W ) and P(W 0 ). Therefore, hd = [X ∩ P(U )] =
[P(W )] + [P(W 0 )] and it follows that [P(W )] 6= [P(W 0 )]. ¤
Exercise 68.3. Determine a complete multiplication table for CH(X) by showing
that
(1) if D is odd then hd+1 = 2ld ;
(2) if D is even and not divisible by 4 then ld2 = 0;
(3) if D is divisible by 4, then ld2 = l0 .
69. ISOMORPHISMS OF QUADRICS 315
Exercise 68.4. Assume that D is even and let ld , ld0 ∈ Ch(X) be two different d-
dimensional subspaces. Let f be the automorphism of Ch(X) induced by a reflection.
Show that f (ld ) = ld0 .
If D is even, an orientation of the quadric is the choice of one of two classes of d-
dimensional linear subspaces in CH(X). We denote this class by ld . An even-dimensional
quadric with an orientation is called oriented.
Proposition 68.5. For any r ≥ 1, the Chow group CH(X r ) is free with basis given
by the external products of the basis elements {hi , li }, i ∈ [0, d], of CH(X).
Proof. The cellular structure on X, constructed in the proof of Proposition 68.1,
together with the calculation of the Chow motive of a projective space (cf. Example
66.5) show by Corollary 66.4 that the motive of X is split. Therefore, the homomorphism
CH(X)⊗r → CH(X r ), given by the external product of cycles is an isomorphism by
Proposition 64.3. ¤
69. Isomorphisms of quadrics

Let ϕ and ψ be two quadratic forms. A similitude between ¡ϕ and¢ ψ (with multiplier
a ∈ F × ) is an isomorphism f : Vϕ → Vψ such that ϕ(v) = aψ f (v) for all v ∈ Vϕ . A
similitude between ϕ and ψ induces an isomorphism of projective spaces P(Vϕ ) → P(Vψ )
∼
∼
and projective quadrics Xϕ → Xψ .
Let i : Xϕ → P(Vϕ ) be the embedding. We consider the locally free sheaves
¡ ¢
OXϕ (s) := i∗ OP(Vϕ ) (s)
over Xϕ for every s ∈ Z.
¡Lemma 69.1.¢ Let ϕ be a ¡nonzero quadratic

¢ form of dimension at least 2. Then
H 0 Xϕ , OXϕ (−1) = 0 and H 0 Xϕ , OXϕ (1) is canonically isomorphic to Vϕ∗ .
¡ ¢ ¡ ¢
Proof. We have H 0 P(Vϕ ), OP(Vϕ ) (−1) = 0, H 0 P(Vϕ ), OP(Vϕ ) (1) ' Vϕ∗ and
¡ ¢
H 1 P(Vϕ ), OP(Vϕ ) (s) = 0 for any s (see [52, Ch. III, Th. 5.1]). The statements follow
from exactness of the cohomology sequence for the short exact sequence
ϕ
0 → OP(Vϕ ) (s − 2) −
→ OP(Vϕ ) (s) → i∗ OXϕ (s) → 0. ¤
∼
Lemma¡ 69.2.¢ Let α : Xϕ → Xψ be an isomorphism of smooth projective quadrics.
∗
Then α OXψ (1) ' OXϕ (1).
Proof. In the case dim ϕ = 2 the sheaves OXϕ (1) and OXψ (1) are free and the
statement is obvious.
We may assume that dim ϕ > 2. As the Picard group of smooth projective varieties
injects under field extensions, we also may assume that both forms are split. We identify
the groups Pic(Xϕ ) and CH1 (Xϕ ). The class of the sheaf OXϕ (1) corresponds to the class
h ∈ CH1 (Xϕ ) of a hyperplane section. It is sufficient to show that α∗ (h) = ±h, since the
class −h cannot occur as the sheaf OXϕ (−1) has no nontrivial global sections by Lemma
69.1.
If dim ϕ > 4 then, by Proposition 68.1, the group CH1 (Xϕ ) is infinite cyclic and
generated by h. Thus α∗ (h) = ±h.
If dim ϕ = 3, then h is twice the generator l0 of the infinite cyclic group CH1 (Xϕ ) and
the result follows in a similar fashion.
Finally, if dim ϕ = 4, then the group CH1 (Xϕ ) is a free abelian group with two
generators l1 and l10 satisfying l1 + l10 = h (cf. the proof of Proposition 68.2). Using
the fact that the pull-back map α : CH∗ (Xϕ ) → CH∗ (Xϕ ) is a ring homomorphism, one
concludes that α∗ (l1 + l10 ) = ±(l1 + l10 ). ¤
Theorem 69.3. Every isomorphism between smooth projective quadrics Xϕ and Xψ
is induced by a similitude between ϕ and ψ.
∼ ¡ ¢
Proof. Let α : Xϕ → Xψ be an isomorphism. By Lemma 69.2, α∗ OXψ (1) '
OXϕ (1). Lemma 69.1 therefore gives an isomorphism of vector spaces
¡ ¢ ∼ ¡ ¢
(69.4) Vψ∗ = H 0 Xψ , OXψ (1) → H 0 Xϕ , OXϕ (1) = Vϕ∗ .
Thus α is given by the induced graded ring isomorphism S • (Vψ∗ ) → S • (Vϕ∗ ) which must
take the ideal (ψ) to (ϕ), i.e., it takes ψ to a multiple of ϕ. In other words, the linear
isomorphism f : Vϕ → Vψ dual to (69.4) is a similitude between ϕ and ψ inducing α. ¤
Corollary 69.5. Let ϕ and ψ be non-degenerate quadratic forms. Then the quadrics
Xϕ and Xψ are isomorphic if and only if ϕ and ψ are similar.
For a quadratic form ϕ all similitudes Vϕ → Vϕ form the group of similitudes GO(ϕ).
For every a ∈ F × , the endomorphism of Vϕ given by the product with a is a simil-
itude. Therefore F × can be identified with a subgroup of GO(ϕ). The factor group
PGO(ϕ) := GO(ϕ)/F × is called the group of projective similitudes. Every projective
similitude induces an automorphism of the quadric Xϕ , so we have a group homomor-
phism PGO(ϕ) → Aut(Xϕ ).
Corollary 69.6. Let ϕ be a non-degenerate quadratic form. Then the map PGO(ϕ) →
Aut(Xϕ ) is an isomorphism.
70. Isotropic quadrics

The motive of a smooth isotropic quadric is computed in terms of a smooth quadric
of smaller dimension in Example 66.7 as follows:
Proposition 70.1. Let X be a smooth isotropic projective quadric given by a quadratic
form ϕ with dim ϕ ≥ 2 and Y a (smooth) projective quadric given by a quadratic form of
dimension dim ϕ − 2 that is Witt equivalent to ϕ (we assume that Y = ∅ if dim X ≤ 1).
Then
M (X) ' Z ⊕ M (Y )(1) ⊕ Z(D)
with D = dim X. In particular,
CH∗ (X) ' CH∗ (Z) ⊕ CH∗−1 (Y ) ⊕ CH∗−D (Z) .
71. THE CHOW GROUP OF DIMENSION 0 CYCLES ON QUADRICS 317
The motivic decomposition of Proposition 70.1 was originally observed by M. Rost.

In most of this book we are interested only in smooth quadrics. Nevertheless, we
make some observations concerning the general (not necessarily smooth) case. Let ϕ be
a quadratic form on a vector space V over F and X the associated projective quadric.
Suppose a rational point x on X is singular. Then ϕ is nonzero and x is determined by
a nonzero vector v ∈ V in the radical of ϕ (cf. proof of Proposition 22.1). Let Y be the
quadric given by the quadratic form on V /F v induced by ϕ, so dim Y = dim X − 1 and
X \ {x} is an affine bundle over Y of rank 1. Since the first homomorphism in the exact
localization sequence
¡ ¢
CH {x} → CH(X) → CH(Y ) → 0
is a¡ split injection
¢ (with the splitting given by the degree homomorphism CH(X) →
CH Spec(F ) ), we have
Lemma 70.2. Let x be a singular rational point of a quadric X and Y be as above.
Then CH0 (X) is (the infinite cyclic group) generated by [x] and CHi (X) is isomorphic to
CHi−1 (Y ) for all i > 0 .
Now assume that ϕ is nonzero and that X = Xϕ has a non-singular rational point
x ∈ X (although X itself may not be smooth). The point x is given by a nonzero isotropic
vector v ∈ V . The projective quadric Y , given by the restriction of ϕ on the orthogonal
complement v ⊥ of v in V , is a closed subscheme of X containing x. Since ϕ 6= 0, the
vector v lies outside of the radical of ϕ, so that v ⊥ is a hyperplane of V . If ϕ restricted
to v ⊥ is zero, then Y is a projective space; otherwise x is a singular point of Y (and
codimX Y = 1). The difference X \ Y is isomorphic to an affine space (cf. proof of
Proposition 22.9). Thus we can apply Theorem 66.2 to X with the one term filtration
Y ⊂ X. Theorem 66.2 and Lemma 70.2 yield (we also include the trivial case of ϕ = 0):
Lemma 70.3. Let x be a non-singular rational point of a quadric X and Y be as
above. Then the push-forward homomorphism CHi (Y ) → CHi (X) is an isomorphism for
any i < dim X. In particular, if dim X > 0, the group CH0 (X) is generated by [x].
Combining Lemma 70.2 with Lemma 70.3 and adding the trivial case of ϕ = 0 yields
Corollary 70.4. Let X be an isotropic (not necessarily smooth) projective quadric
X. Then the degree homomorphism deg : CH0 (X) → Z is an isomorphism unless ϕ is a
hyperbolic plane, i.e., X is a disjoint union of two copies of Spec F .
71. The Chow group of dimension 0 cycles on quadrics

Recall that for every p = 0, 1, . . . , n, the group CHp (PnF ) is infinite cyclic generated
by the class hp where h ∈ CH1 (PnF ) is the class of a hyperplane in PnF (cf. Example
57.22). Thus for every p = 0, 1, . . . , n and α ∈ CHp (PnF ), we have α = mhp for a uniquely
determined integer m. We call m the degree of α and write m = deg(α). We have
deg(αβ) = deg(α) deg(β) for all homogeneous cycles α ∈ CHp (PnF ) and β ∈ CHq (PnF )
satisfying p, q ≥ 0 and p + q ≤ n.
If Z is a closed subvariety of PnF , we define the degree deg(Z) of Z as deg([Z]).
Lemma 71.1. Let x ∈ PnF be a closed point of degree d > 1 such that the field extension
F (x)/F is simple (i.e., generated by one element). Then there is a morphism f : P1 → Pn
with image a curve C satisfying x ∈ C and deg(C) < d.
Proof. Let u be a generator of the field extension F (x)/F . We can write the homoge-
neous coordinates si of x in the form si = fi (u), i ∈ [0, n], where fi are polynomials over F
of degree less than d. Let k be the largest degree of the fi and set Fi (t0 , t1 ) = tk1 fi (t0 /t1 ).
The polynomials Fi are all homogeneous of degree k < d. We may assume that all
the Fi are relatively prime (by dividing out the gcd of the Fi ). Consider the morphism
f : P1F → PnF given by the polynomials Fi and let C be the image of f . Note that C
contains x and C(F ) 6= ∅. In particular, the map f is not constant. Therefore C is a
closed curve in PnF . We have f∗ ([P1 ]) = r[C] for some r ≥ 1.
Choose an index i such that Fi is a nonzero polynomial and consider the hyperplane
H in PnF given by si = 0. The subscheme f −1 (H) ⊂ P1F is given by Fi (t0 , t1 ) = 0, so
f −1 (H) is a 0-dimensional subscheme of degree k = deg Fi . Hence H has a proper inverse
image with respect to f . By Proposition 57.18, we have f ∗ (h) = mp, where p is the class
of a point in P1F and 1 ≤ m ≤ k < d. It follows from Proposition 56.9 that
¡ ¢ ¡ ¢
h · r[C] = h · f∗ [P1 ] = f∗ f ∗ (h) = f∗ (mp) = mhn .
Hence deg(C) = m/r ≤ m < d. ¤
We follow [68, Th. 3.2] in the proof of the next result.
Theorem 71.2. Let X be an anisotropic (not necessarily smooth) quadric over F and
let x0 ∈ X be a closed point of degree 2. Then for every closed point x ∈ X, we have
[x] = a[x0 ] in CH0 (X) for some a ∈ Z.
Proof. We proceed by induction on d = deg x. Suppose first that there are no
intermediate fields between F and F (x). In particular, the field extension F (x)/F is
simple. The quadric X is a hypersurface in the projective space PnF for some n. By
Lemma 71.1, there is an integral closed curve C ⊂ PnF of degree less that d with C(F ) 6= ∅
and x ∈ C.
Let g : X → PnF be the closed embedding. Since X is anisotropic and C(F ) 6= ∅,
C is not contained in X. Therefore, C has proper inverse image with respect to g. As
x ∈ C ∩ X, by Proposition 57.18,
g F ([C]) = [x] + α
in CH0 (X), where α is a non-negative zero-dimensional cycle on X. By Proposition 56.11,
we have ¡ ¢ ¡ ¢
deg g F ([C]) = deg g∗ ◦ g F ([C]) = deg([X] · [C]) = 2 deg(C),
hence
deg α = 2 deg C − deg x = 2 deg C − d < d.
Thus the cycle α is supported on closed points of degree less that d. By the induction
hypothesis, α = b[x0 ] in CH0 (X) for some b ∈ Z. We also have [C] = c[L] in CH0 (PnF )
where L is a line in PnF satisfying x0 ∈ L and c = deg C. Since L ∩ X = {x0 }, by Corollary
57.19, we have g F ([L]) = [x0 ]. Therefore,
[x] = g F ([C]) − α = cg F ([L]) − b[x0 ] = (c − b)[x0 ].
72. THE REDUCED CHOW GROUP 319
Now suppose that there is a proper intermediate field L between F and E = F (x). Let
f denote the natural morphism XL → X. The morphism Spec E → X induced by x and
the inclusion of L into E defines a closed point x0 ∈ XL with f (x0 ) = x and F (x0 ) = E.
It follows that f∗ ([x0 ]) = [x].
Consider two cases:
¡ ¢
Case 1: XL is isotropic. Let¡y ∈¢ XL be a rational point. As deg f∗ [y] = [L : F ] < d, by
the induction hypothesis, f∗ [y] ∈ Z · [x0 ]. By Corollary 70.4,
¡ 0 ¢ we know that CH0 (XL ) is
generated by the classes of rational points, hence [x] = f∗ [x ] ∈ Z · [x0 ].
Case 2: XL is anisotropic. Applying the induction
£ hypothesis
¤ to the quadric XL and the
point x of degree [E : L] < d, we have [x ] = b (x0 )L for some b ∈ Z. Hence
0 0
¡ ¢
[x] = f∗ [x0 ] = bc[x0 ],
where c = [L : F ]. ¤
We therefore obtain another proof of Springer’s Theorem 18.5.
Corollary 71.3 (Springer’s Theorem). If X is an anisotropic quadric, the image of
the degree homomorphism deg : CH0 (X) → Z is equal to 2Z, i.e., the degree of a finite
field extension L/F with XL isotropic, is even.
In the case char F 6= 2, the following important statement was proven in [70, Prop.
2.6] and by Swan in [135]. The case of characteristic 2 was considered by Totaro in [137].
Corollary 71.4. For every anisotropic quadric X, the degree homomorphism deg :
CH0 (X) → Z is injective.
72. The reduced Chow group

Let X be an arbitrary smooth projective quadric. We write CH(X̄ r ) for CH(XEr ),
where E is a field extension of F such that the quadric XE is split. Note that for any field L
containing E, the change of field homomorphism CH(XEr ) → CH(XLr ) of Example 49.14 is
an isomorphism; therefore for any field extension E 0 /F with split XE 0 , the groups CH(XEr )
and CH(XEr 0 ) are canonically isomorphic, hence CH(X̄ r ) can be defined invariantly as the
colimit of the groups CH(XLr ), where L runs over all field extensions of F .
If D is even, an orientation of the quadric is the choice of one of two classes of d-
dimensional linear subspaces in CH(X̄). An even-dimensional quadric with an orientation
is called oriented.
The reduced Chow group CH(X r ) is defined as the image of the change of field homo-
morphism CH(X r ) → CH(X̄ r ).
We say that an element of CH(X̄ r ) is rational if it lies in the subgroup CH(X r ) ⊂
CH(X̄ r ). More generally, for a field extension L/F , the elements of the subgroup CH(XLr ) ⊂
CH(X̄ r ) are called L-rational.
Replacing the integral Chow group by the Chow group modulo 2 in the above defini-
tions, we get the modulo 2 reduced Chow group Ch(X r ) ⊂ Ch(X̄ r ) and the corresponding
notion of (L-)rational cycles modulo 2.
Abusing notation, we shall often call elements of a Chow group cycles. The basis
described in Proposition 68.5 will be called a basis for CH(X̄ r ) and its elements basis
elements or basic cycles. Similarly, this basis modulo 2 will be called a basis for Ch(X̄ r )
and its elements basis elements or basic cycles. We use the same notation for the basis
elements of CH(X̄) and for their reductions modulo 2. The decomposition of an element
α ∈ Ch(X̄ r ) will always mean its representation as a sum of basic cycles. We say that a
basis cycle β is contained in the decomposition of α (or simply “is contained in α”), if β
is a summand of the decomposition. More generally, for two cycles α0 , α ∈ Ch(X̄ r ), we
say that α0 is contained in α or that α0 is a subcycle of α (notation: α0 ⊂ α), if every basis
element contained in α0 is also contained in α.
A basis element of Ch(X̄ r ) is called non-essential, if it is an external product of
(internal) powers of h (including h0 = 1 = [X̄]); the other basis elements are called
essential. An element of Ch(X̄ r ) that is a sum of non-essential basis elements, is called
non-essential as well. Note that all non-essential elements are rational since h is rational.
An element of Ch(X̄ r ) that is a sum of essential basis elements, is called essential as well.
(The zero cycle is the only element which is essential and non-essential simultaneously).
The group Ch(X̄ r ) is a direct sum of the subgroup of non-essential elements and the
subgroup of essential elements. We call the essential component of an element α ∈ Ch(X̄ r )
the essence of α. Clearly, the essence of a rational element is rational.
The group Ch(X) is easy to compute. First of all, by Springer’s theorem (Corollary
71.3), one has
Lemma 72.1. If the quadric X is anisotropic (i.e., X(F ) = ∅), then the element
l0 ∈ Ch(X̄) is not rational.
Corollary 72.2. If X is anisotropic, the group Ch(X) is generated by the non-
essential basis elements.
Proof. If the decomposition of an element α ∈ Ch(X) contains an essential basis
element li for some i 6= D/2, then li ∈ Ch(X) because li is the i-dimensional homogeneous
component of α (and Ch(X) is a graded subring of Ch(X̄)). If the decomposition of an
element α ∈ Ch(X) contains the essential basis element li for i = D/2 then D/2 = d,
and the d-dimensional homogeneous component of α is either ld or ld + hd so we still have
li ∈ Ch(X). It follows that l0 = li · hi ∈ Ch(X), contradicting Lemma 72.1. ¤
Let V be the underlying vector space of ϕ and W ⊂ V a totally isotropic subspace of
dimension a ≤ d. Let Y be the projective quadric of the quadratic form ψ : W ⊥ /W → F
induced by ϕ. Then ψ is non-degenerate, Witt-equivalent to ϕ, dim ψ = dim ϕ − 2a, and
dim Y = dim X − 2a. Let Z ⊂ Y × X be the closed scheme of the pairs (y, x) satisfying
the condition x ∈ p−1 (y), where p is the projection W ⊥ → W ⊥ /W . Note that the
pr
composition Z ,→ Y × X →Y Y is an a-dimensional projective bundle; in particular, Z is
equidimensional (and Z is a variety if Y is) of dimension dim Z = dim Y + a = dim X − a.
Its class α = [Z] ∈ CH(Y × X) is called the incidence
¡ ¢ correspondence.
We first note that the inverse image pr −1
X P(W ) of the closed subvariety P(W ) ⊂ X
under the projection pr X : Y × X → X is contained in Z ¡with complement¢ a dense open
subscheme of Z mapping under pr X isomorphically onto P(W ⊥ ) ∩ X \ P(W ).
We let hi = 0 = li for any negative integer i.
72. THE REDUCED CHOW GROUP 321
Lemma 72.3. Let α be the incidence correspondence in CH(Y × X).

(1) For any i = 0, . . . , d − a, the homomorphism α∗ : CH(Ȳ ) → CH(X̄) takes hi to
hi+a and li to li+a .
(2) For any i = 0, . . . , d, the homomorphism α∗ : CH(X̄) → CH(Ȳ ) takes hi to hi−a
and li to li−a . (In the case of even D, the two formulae involving ld are true for
an appropriate choice of orientations of X and of Y .)
Proof. For an arbitrary i ∈ [0, d − a], let L ⊂ W ⊥ /W be a totally isotropic lin-
ear subspace ¡ of dimension
¡ ¢ i ¢+ 1. −1 ¡ li ¢= [P(L)] ∈ CH(Y ). Since
Then ¡ the ¢dense open
subscheme pr −1 Y P(L) ∩ Z \ pr X P (W ) of the intersection pr Y P(L) ∩ Z maps
−1
under pr X isomorphically onto P(p (L)) \ P(W ), we have (using Proposition 57.20):
−1
α∗ (li ) = [P(p−1 (L))] = li+a ∈ CH(X). Similarly, for any linear subspace H ⊂ W ⊥ /W of
codimension i, the element hi ∈ CH(Y ) is the class of the intersection P(H) ∩ Y , maps
under α∗ to the class of [P(p−1 (H)) ∩ X] which equals hi+a .
To prove the statements on α∗ for an arbitrary i ∈ [a, d], let L ⊂ V be an (i + 1)-
dimensional totally isotropic subspace satisfying dim(L ∩ W ⊥ ) = dim L − a and L ∩
W = 0. (The second condition is, in fact, a consequence of the first one.) Then li =
[P(L)] ¡∈ CH(X) and the intersection
¢ X (P(L)) ∩ Z maps under pr Y isomorphically
pr −1
onto P ((L ∩ W ) + W )/W ; consequently, α∗ (li ) = li−a . Similarly, if H ⊂ V is a linear
⊥
subspace of codimension i satisfying dim(H ∩ W ⊥ ) = ¡dim H − a ¢and H ∩ W = 0, then

hi = [P(H) ∩ X] ∈ CH(X) ¡ and the intersection
¢ X P(H) ∩ X ∩ Z maps under pr Y
pr −1
isomorphically onto P ((H ∩ W ) + W )/W ∩ Y ; consequently, α∗ (hi ) = hi−a .
⊥
¤
Corollary 72.4. Assume that X is isotropic but not split and set a = i0 (X). Let
X0 be the projective quadric given by an anisotropic quadratic form Witt-equivalent to ϕ
(so dim X0 = D − 2a). Then
(1) The group ChD−a (X × X0 ) contains a correspondence pr such that the induced
homomorphism pr ∗ : Ch(X̄) → Ch(X̄0 ) takes hi to hi−a and li to li−a for i =
0, . . . , d.
(2) The group ChD−a (X0 × X) contains a correspondence in such that the induced
homomorphism in ∗ : Ch(X̄0 ) → Ch(X̄) takes hi to hi+a and li to li+a for i =
0, . . . , d − a.
Remark 72.5. Note that the homomorphisms in ∗ and pr ∗ of Corollary 72.4 map
rational ¡cycles to¢ rational cycles. Since the composite pr ∗ ◦ in ∗ is the identity, it follows
that pr ∗ Ch(X) = Ch(X0 ). More generally, for any r ≥ 1 the homomorphisms
in r∗ : Ch(X̄0r ) → Ch(X̄ r ) and pr r∗ : Ch(X̄ r ) → Ch(X̄0r ) ,
induced by the r-th tensor powers in r ∈ Ch(X0r × X r ) and pr r ∈ Ch(X r × X0r ) of the
correspondences in and ¡ pr , map¢ rational rcycles to rational cycles and satisfy the relations
r r r r
pr ∗ ◦ in ∗ = id and pr ∗ Ch(X ) = Ch(X0 ).
We obtain now the following extension of Lemma 72.1.
Corollary 72.6. Let X be an arbitrary quadric and i any integer. Then li ∈ Ch(X)
if and only if i0 (X) > i.
Proof. The “if” part of the statement is trivial. We prove the “only if” part by
induction on i. The case i = 0 is Lemma 72.1.
We may assume that i > 0 and li ∈ Ch(X). Since li · h = li−1 , the element li−1 is
also rational. Therefore i0 (X) ≥ i by the induction hypothesis. If i0 (X) = i the image
of li ∈ Ch(X) under the map pr ∗ : Ch(X̄) → Ch(X̄0 ) of Corollary 72.4 equals l0 and is
rational. Therefore, by Lemma 72.1, the quadric X0 is isotropic, a contradiction. ¤
The following observation is crucial:
Theorem 72.7. The absolute and relative higher Witt indices of a non-degenerate
quadratic form ϕ are determined by the group
L
Ch(X ∗ ) = Ch(X r ) .
r≥1
Proof. We first note that the group Ch(X) determines i0 (ϕ) by Corollary 72.6.
Let X0 be as in Corollary 72.4. By Corollary 72.4 and Remark 72.5, the group Ch(X0∗ )
is recovered as the image of the group Ch(X ∗ ) under the homomorphism Ch(X̄ ∗ ) →
Ch(X̄0∗ ) induced by the tensor powers of the correspondence pr .
Let F1 be the first field in the¡ generic
¢ splitting tower of ϕ. The pull-back ho-
momorphism g1∗ : Ch(X0r ) → Ch (X0 )r−1 F1 with respect to the morphism of schemes
r−1 r
g1 : (X0 )F1 → X0 given by the generic point of the first factor of X0r , is surjective
¡ ¢
(cf. Corollary 57.10). It induces an epimorphism Ch(X0r ) ³ Ch (X0 )r−1 F1 , which is
the restriction of the epimorphism Ch(X̄0r ) ³ Ch(X̄0r−1 ) mapping each basis element of
the form h0 × β, β ∈ Ch(X̄0r−1 ), to β and
¡ killing
¢ all other basis elements. Therefore, the
∗ ∗
group Ch(X0 ) determines the group Ch (X0 )F1 , and we finish by induction on the height
h(ϕ). ¤
Remark 72.8. The proof of Theorem 72.7 shows that the statement of Theorem 72.7
can be made more precise in the following way: If for some q ∈ [0, h] the absolute Witt
indices j0 , . . . , jq−1 are already known, then one determines jq by the formula
jq = max{j | the product hj0 × hj1 × · · · × hjq−1 × lj−1 is contained in a rational cycle} .
73. Cycles on X 2
In this section, we study the groups Chi (X 2 ) for i ≥ D. After Lemma 73.2, we shall
assume that X is anisotropic.
Most results of this section are simplified versions of original results on integral motives
of quadrics due to Vishik in [140].
Lemma 73.1. The sum
P
d
∆= (hi × li + li × hi ) ∈ Ch(X̄ 2 )
i=0
is always rational.
Proof. Either the composition with correspondence ∆ or the composition with the
correspondence ∆ + hd × hd (depending on whether ld2 is zero or not) induces the identity
endomorphism of Ch(X̄ 2 ). Therefore, this correspondence is the class of the diagonal
which is rational. ¤
73. CYCLES ON X 2 323
Lemma 73.2. Suppose for some i ∈ [1, d] at least one of the basis elements ld × li or
li × ld of the group Ch(X̄ 2 ) appears in the decomposition of a rational cycle. Then X is
hyperbolic.
Proof. Let α be a cycle in Chi+d (X 2 ) containing li × ld or ld × li . Replacing α by its
transpose if necessary we may assume that li × ld ∈ α. If β ∈ CHi+d (X̄ 2 ) is a basic cycle,

 ld if β = li × ld
i
β∗ (h ) = hd if β = li × hd

0 otherwise,
hence α∗ (hi ) is equal to ld or hd + ld . As α∗ (hi ) is rational, the cycle ld is rational and X
is hyperbolic by Corollary 72.6. ¤
We assume now that X is anisotropic throughout the rest of this section.
Let α1 , α2 ∈ Ch∗ (X 2 ). The intersection α1 ∩ α2 denotes the sum of the basic cycles
contained simultaneously in α1 and in α2 .
L
Lemma 73.3. If α1 , α2 ∈ i≥0 ChD+i (X 2 ) then the cycle α1 ∩ α2 is rational.
Proof. Clearly, we may assume that α1 and α2 are homogeneous of the same dimen-
sion D + i and do not contain any non-essential basis element. Using Lemma 73.2 we see
that ¡the intersection
¢ α1 ∩ α2 is the essence of the composite of rational correspondences
0 i
α2 ◦ α1 · (h × h ) and hence is rational. ¤
L Definition 73.4. We write Che(X 2 ) for the group of essential rational elements in
2
i≥D Chi (X̄ ).
Definition 73.5. A non-zero element of Che(X 2 ) is called minimal if it does not

contain any proper rational subcycle.
Note that a minimal cycle is always homogeneous.
Proposition 73.6. Let X be a smooth anisotropic quadric. Then
(1) The minimal cycles form a basis of the group Che(X 2 ).
(2) Two different minimal cycles intersect trivially.
P
(3) The sum of the minimal cycles of dimension D is equal to the sum di=0 hi × li +
li × hi of all D-dimensional essential basis elements (excluding ld × ld in the case
of even D).
Proof. The first two statements of Proposition 73.6 follow from Lemma 73.3. The
last statement follows from the previous ones together with Lemma 73.1. ¤
Definition 73.7. Let α be an element of ChD+r (X̄ 2 ) for some r ≥ 0. For every
i ∈ [0, r], the products α · (h0 × hi ), α · (h1 × hi−1 ), . . . , α · (hi × h0 ) will be called the (i-th
order) derivatives of α.
Note that all the derivatives of a rational cycle are also rational.
Lemma 73.8. (1) A derivative of an essential basis element β ∈ CheD+r (X̄ 2 ) is
an essential basis element.
(2) For any r ≥ 0, non-negative integers i1 , j1 , i2 , j2 satisfying i1 + j1 ≤ r, i2 + j2 ≤ r,

and non-zero essential cycle β ∈ CheD+r (X̄ 2 ), the two derivatives β · (hi1 ×
hj1 ) and β · (hi2 × hj2 ) of β coincide only if i1 = i2 and j1 = j2 .
(3) For any r ≥ 0, non-negative integers i, j with i + j ≤ r, and non-zero essential
cycles β1 , β2 ∈ CheD+r (X̄ 2 ), the derivatives β1 · (hi × hj ) and β2 · (hi × hj ) of
β1 and β2 coincide only if β1 = β2 .
Proof. (1): If β is an essential basis element of CheD+r (X̄ 2 ) for some r > 0, then
up to transposition, β = hi × li+r for some i ∈ [0, d − r]. An arbitrary derivative of β is
equal to β · (hj1 × hj2 ) = hi+j1 × li+r−j2 for some j1 , j2 ≥ 0 such that j1 + j2 ≤ r. It follows
that the integers i + j1 and i + r − j2 are in the interval [0, d]; therefore, hi+j1 × li+r−j2
is an essential basis element.
Statement (2) and (3) are left to the reader. ¤
Remark L 73.9. To 2visualize the above, it is convenient to think of the essential basic
cycles in i≥D Chi (X̄ ) (with lD/2 × lD/2 excluded by Lemma 73.2) as points of two
“pyramids”. For example, if D = 8 or D = 9, we write
∗ ∗
∗ ∗ ∗ ∗
∗ ∗ ∗ ∗ ∗ ∗
∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗
∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗
If we count the rows of the pyramids from the bottom starting with 0, the top row has
number d, and for every r = 0, . . . , d, the rth row of the left pyramid represents the
essential basis elements hi × lr+i , i = 0, 1, . . . , d − r of ChD+r (X̄ 2 ), while the rth row of
the right pyramid represents the essential basis elements lr+i ×hi , i = d−r, d−r −1, . . . , 0
(so that the basis elements of each row are ordered by the codimension of the first factor).
For any α ∈ Ch(X̄ 2 ), we fill in the pyramids by putting a mark in the points repre-
senting basis elements contained in the decomposition of α; the pictures thus obtained is
the diagram of α. If α is homogeneous, the marked points (if any) lie in the same row.
It is now easy to interpret the derivatives of α if α is homogeneous of dimension ≥ D:
the diagram of an i-th order derivative is a parallel transfer of the marked points of the
diagram of α moving them i rows lower. In particular, the diagram of every derivative of
such an α has the same number of marked points as the diagram of α (cf. Lemma 73.8).
The diagrams of the two different derivatives of the same order are shifts (to the right or
to the left) of each other.
Example 73.10. Let D = 8 or D = 9. Let α ∈ ChD+1 (X̄ 2 ) be the essential cycle

α = h0 × l1 + h2 × l3 + l3 × h2 . Then the diagram is
∗ ∗
∗ ∗ ∗ ∗
∗ ∗ ∗ ∗ ∗ ∗
• ∗ • ∗ ∗ • ∗ ∗
∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗
Then α has precisely two first order derivatives. They are given by α · (h0 × h1 ) =
h0 × l0 + h2 × l2 + l3 × h3 and α · (h1 × h0 ) = h1 × l1 + h3 × l3 + l2 × h2 . Their diagrams
are as follows:
∗ ∗
∗ ∗ ∗ ∗
∗ ∗ ∗ ∗ ∗ ∗
∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗
• ∗ • ∗ ∗ ∗ • ∗ ∗ ∗
∗ ∗
∗ ∗ ∗ ∗
∗ ∗ ∗ ∗ ∗ ∗
∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗
∗ • ∗ • ∗ ∗ ∗ • ∗ ∗
Lemma 73.11. Let α ∈ Che(X 2 ). Then the following conditions are equivalent:
(1) α is minimal.
(2) All derivatives of α are minimal.
(3) At least one derivative of α is minimal.
Proof. Derivatives of a proper subcycle of α are proper subcycles of the derivatives
of α; therefore, (3) ⇒ (1).
In order to show that (1) ⇒ (2), it suffices to show that the two first order derivatives
α · (h0 × h1 ) and α · (h1 × h0 ) of a minimal cycle α are also minimal. If not, possibly
replacing α by its transposition, we reduce to the case where the derivative α · (h0 × h1 )
of a minimal α is not minimal. It follows that the cycle α · (h0 × hi ) with i = dim α − D
is also not minimal. Let α0 be its proper subcycle. Taking the essence of the composite
α ◦ α0 , we get a proper subcycle of α, a contradiction. ¤
Corollary 73.12. The derivatives of a minimal cycle are disjoint.
Proof. The derivatives of a minimal cycle are minimal by Lemma 73.11 and pairwise
different by Lemma 73.8. As two different minimal cycles are disjoint by Lemma 73.3,
the result follows. ¤
We recall the notation of §25. If F0 = F, F1 , . . . , Fh is the generic splitting tower of
ϕ with h = h(ϕ) the height of ϕ, and ϕi = (ϕFi )an for i ≥ 0, we let Xi = Xϕi be the
projective quadric over Fi given by ϕi . Then ir = ir (ϕ) = ir (X) is the rth relative and
jk = jk (ϕ) = jk (X) the rth absolute higher Witt index of ϕ, r ∈ [0, h]. We will also call
these numbers the relative and the absolute Witt indices of X respectively.
Lemma 73.13. Let i, j be integers in the interval [0, d] satisfying i < jq ≤ j for some
q ∈ [1, h) then no element in Ch(X 2 ) contains either hi × lj or lj × hi .
Proof. Let i, j be integers of the interval [0, d] such that hi × lj or lj × hi appears
in the decomposition of some α ∈ Ch(X 2 ). Replacing α by its transpose if necessary we
may assume that hi × lj ∈ α. Replacing α by its homogeneous component containing
hi × lj , we reduce to the case that α is homogeneous.
Suppose q be an integer in [1, h) satisfying i < jq . It suffices to show that j < jq as
well.
Let L be a field extension of F with i0 (XL ) = jq (e.g., L = Fq ). The cycles α and li
are both L-rational. Therefore, so is the cycle α∗ (li ) = lj . It follows by Corollary 72.6
that j < jq . ¤
Remark 73.14. In order to “see” the statement of Lemma 73.13, it is helpful to mark
by a ∗ only those essential basis elements that are not “forbidden” by this lemma in the
pyramids of basic cycles drawn in Remark 73.9 and to mark by a ◦ the remaining points
of the pyramids. We will get isosceles triangles based on the lower row of these pyramids.
For example, if X is a 34-dimensional quadric with relative higher Witt indices 4, 2, 4, 8,
the picture looks as follows:
◦ ◦◦◦◦◦ ◦ ◦◦◦◦◦
◦ ◦◦◦◦◦◦◦◦◦◦◦◦◦◦ ◦ ◦◦◦◦◦◦◦◦◦◦◦◦◦◦
◦ ◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦ ◦ ◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦
◦ ◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦∗◦∗∗ ∗ ∗∗◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦
∗ ◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦∗◦◦◦◦◦◦◦◦◦◦◦◦∗∗∗∗∗∗∗∗∗∗∗∗ ∗ ∗∗∗∗∗∗∗∗∗∗◦∗◦◦◦◦◦◦◦◦◦◦∗◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦∗
∗ ∗ ◦ ◦ ◦ ◦ ∗ ∗ ◦ ◦ ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ ◦ ◦ ∗ ∗ ◦ ◦ ◦ ◦ ∗ ∗
∗∗ ∗ ∗ ∗∗ ∗ ◦ ∗∗ ∗ ◦ ∗ ∗ ∗∗ ∗ ∗ ∗◦ ∗ ∗ ∗ ∗ ∗∗ ∗ ∗ ∗ ∗ ∗∗ ∗ ∗ ∗ ∗ ∗ ∗∗ ∗ ∗ ∗∗ ∗ ∗ ∗ ∗ ∗∗ ∗ ◦ ∗ ∗∗ ∗ ∗ ∗ ∗◦ ∗ ∗ ∗ ◦ ∗∗ ∗ ∗ ∗∗ ∗
Definition 73.15. The triangles in the pyramid of Remark 73.14 will be called shell
triangles. The shell triangles in the left pyramid are numbered from the left starting with
1. The shell triangles in the right pyramid are numbered from the right starting by 1 as
well (so that the symmetric triangles have the same number; for any q ∈ [1, h], the bases
of the q-th triangles have (each) iq points). The rows of the shell triangles are numbered
from below starting with 0. The points of rows of the shell triangles (of the left ones as
well as of the right ones) are numbered from the left starting with 1.
L
Lemma 73.16. For every rational cycle α ∈ i≥D Chi (X 2 ), the number of essential
basic cycles contained in α is even (i.e., the number of the marked points in the diagram
of α is even).
Proof. We may assume that α is homogeneous, say, α ∈ ChD+k (X 2 ), k ≥ 0. We
may also assume that k ≤ d as there are no essential basic cycles of dimension > D + d.
Let n be the number of essential basic cycles contained in α. The pull-back δ ∗ (α) of α
with respect to the diagonal δ : X → X 2 produces n · lk ∈ Ch(X). It follows by Corollary
72.2 that n is even. ¤
Lemma 73.17. Let α ∈ Ch(X 2 ) be a cycle containing the top of a qth shell triangle
for some q ∈ [1, h]. Then α also contains the top of the other qth shell triangle.
Proof. We may assume that α contains the top of left qth shell triangle. Replacing
F by the field Fq−1 in the generic splitting tower of F , X by Xq−1 , and α by pr 2∗ (α),
where pr ∈ Ch(XFq−1 × Xq−1 ) is the correspondence of Corollary 72.4, we may assume
that q = 1.
Replacing α by its homogeneous component containing the top of the left 1st shell
triangle β = h0 × lj1 −1 , we may assume that α is homogeneous.
Suppose that the transpose of β is not contained in α. By Lemma 73.13, the element
α does not contain any essential basic cycles having hi with 0 < i < i1 as a factor. Since
α 6= β by Lemma 73.16, we have h > 1. Moreover, the number of essential basis elements
contained in α and the number of essential basis elements contained in pr 2∗ (α) ∈ Ch(X12 )
differ by 1. In particular, these two numbers have different parity. However, the number
of the essential basis elements contained in α is even by Lemma 73.16. By the same
lemma, the number of essential basis elements contained in pr 2∗ (α) is even, too. ¤
Definition 73.18. A minimal cycle α ∈ Che(X 2 ) is called primordial if it is not a
derivative of positive order of another rational cycle.
Lemma 73.19. Let α ∈ Ch(X 2 ) be a minimal cycle containing the top of a qth shell
triangle for some q ∈ [1, h]. Then α is symmetric and primordial.
Proof. The cycle α ∩ t(α), where t(α) is the transpose of α (where the intersection
of cycles is defined in Lemma 73.3) is symmetric, rational by Lemma 73.3, contained in
α, and, by Lemma 73.17, still contains the tops hjq−1 × ljq −1 and ljq −1 × hjq−1 of both qth
shell triangles. Therefore, it coincides with α by the minimality of α.
It is easy to “see” that α is primordial by looking at the picture in Remark 73.14.
Nevertheless, we prove it. If there exists a rational cycle β 6= α such that α is a derivative
of β, then there exists a rational cycle β 0 such that α is an order one derivative of β 0 , i.e.,
α = β 0 · (h0 × h1 ) or α = β 0 · (h1 × h0 ). In the first case, β 0 would contain the basic cycle
hjq−1 × ljq while in the second case β 0 would contain hjq−1 −1 × ljq −1 . However, neither of
these two cases is possible by Lemma 73.13. ¤
It is easy to see that a cycle α satisfying the hypothesis of Lemma 73.19 with q = 1
exists:
Lemma 73.20. There exists a cycle in ChD+i1 −1 (X 2 ) containing the top h0 × li1 −1 of
the 1st left shell triangle.
Proof. If D = 0, this follows by Lemma 73.1. So assume D > 0. Consider the pull-
back homomorphism Ch(X 2 ) ³ Ch(XF (X) ) with respect to the morphism XF (X) → X 2
produced by the generic point of the first factor of X 2 . By Corollary 57.10, this is an
epimorphism. It is also a restriction of the homomorphism Ch(X̄ 2 ) → Ch(X̄) mapping
each basis element of the type h0 × li to li and vanishing on all other basis elements.
Therefore an arbitrary preimage of li1 −1 ∈ Ch(XF (X) ) under the surjection Ch(X 2 ) ³
Ch(XF (X) ) contains h0 × li1 −1 . ¤
Lemma 73.21. Let ρ ∈ ChD (X 2 ), q ∈ [1, h], and i ∈ [1, iq ]. Then the element
hjq−1 +i−1 × ljq−1 +i−1 is contained in ρ if and only if the element ljq −i × hjq −i is contained
in ρ.
Proof. Clearly, it suffices to prove Lemma 73.21 for q = 1. By Lemma 73.20, the
basis element h0 × li1 −1 is contained in a rational cycle. Let α be the minimal cycle
containing h0 × li1 −1 . By Lemma 73.17, the cycle α also contains li1 −1 × h0 . Therefore, the
derivative α · (hi−1 × hi1 −i ) contains both hi−1 × li−1 and li1 −i × hi1 −i . Since the derivative
of a minimal cycle is minimal by Lemma 73.11, the lemma follows by Lemma 73.3. ¤
In the language of diagrams, the statement of Lemma 73.21 means that the ith point
of the base of the qth left shell triangle in the diagram of ρ is marked if and only if the
ith point of the base of the qth right shell triangle is marked.
Definition 73.22. The symmetric shell triangles (i.e., both qth shell triangles for
some q) are called dual. Two points are called dual, if one of them is in a left shell
triangle, while the other one is in the same row of the dual right shell triangle and has
the same number as the first point.
Corollary 73.23. In the diagram of an element of Ch(X 2 ), any two dual points are
simultaneously marked or not marked.
Proof. Let k be the number of the row containing two given dual points. The case of
k = 0 is treated in Lemma 73.21 (while Lemma 73.17 treats the case of “locally maximal”
k). The case of an arbitrary k is reduced to the case of k = 0 by taking a k-th order
derivative of α. ¤
Remark 73.24. By Corollary 73.23, it follows that the diagram of a cycle in Ch(X 2 ) is
determined by one (left or right) half of itself. From now on, we refer to left shell triangles
simply as shell triangles. Note also that the transposition of a cycle acts symmetrically
about the vertical axis on each shell triangle.
The following proposition generalizes Lemma 73.20.
Proposition 73.25. Let f : Ch(X 2 ) → [1, h] be the map that assigns to each γ ∈
Ch(X 2 ) the integer q ∈ [1, h] such that the diagram of γ has a point in the ¡ qth shell
¢
triangle and has no points in the ith shell triangles with i < q. For any q ∈ f Ch(X 2 ) ,
there exists an element α ∈ Ch(X 2 ) with f (α) = q so that α contains the top of the qth
shell triangle.
Proof. We induct on q. If q = 1, the condition of Proposition 73.25 is automatically
satisfied by Lemma 73.1 and the result follows by Lemma 73.20. So we may assume that
q > 1.
Let γ be an element of Ch(X 2 ) with f (γ) = q. Replacing γ by its appropriate homo-
geneous component, we may assume that γ is homogeneous. Replacing this homogeneous
γ by any one of its maximal order derivatives, we may further assume that γ ∈ ChD (X 2 ).
Let i be the smallest integer such that hjq−1 +i × ljq−1 +i ∈ γ. We first prove that the
group Ch(X 2 ) contains a cycle γ 0 satisfying f (γ 0 ) = q with γ 0 containing hjq−1 +i × ljq −1 .
(This is the point on the right side of the qth shell triangle such that the line connecting
it with hjq−1 +i × ljq −1+i is parallel to the left side of the shell triangle. If i = 0 then we
can take α = γ 0 and finish the proof.)
Let
pr 2∗ : Ch(XF2 (X) ) → Ch(X12 ) and in 2∗ : Ch(X12 ) → Ch(XF2 (X) )
be the homomorphisms of Remark 72.5. Applying the induction hypothesis to the quadric
X1 and the cycle pr 2∗ (γ) ∈ Ch(X12 ), we get a homogeneous cycle in ChD+iq −1 (XF2 (X) )
containing hjq−1 ×ljq −1 . Multiplying it by hi ×h0 , we get a homogeneous cycle in Ch(XF2 (X) )
containing hjq−1 +i × ljq −1 . Note that the quadric XF (X) is not hyperbolic (since h ≥ q > 1)
and therefore, by Lemma 73.2, the basis element ld × ld is not contained in this cycle.
Therefore, the group Ch(X 3 ) contains a homogeneous cycle µ containing h0 ×hjq−1 +i ×ljq −1
(and not containing h0 × ld × ld ). View µ as a correspondence of the middle factor of X 3
into the product of the two outer factors. Composing it with γ and taking the pull-back
with respect to the partial diagonal map δ : X 2 → X 3 , (x1 , x2 ) 7→ (x1 , x1 , x2 ), we get the
¡ ¢
required cycle γ 0 (more accurately, γ 0 = δ t12 (µ) ◦ γ , where t12 is the automorphism of
∗
Ch(X 3 ) given by the transposition of the first two factors of X 3 ).

The highest order derivative γ 0 · (hiq −1−i × h0 ) of γ 0 contains hjq −1 × ljq −1 , the last point
of the base of the qth shell triangle. Therefore, the transpose t(γ 0 ) contains the first point
hjq−1 × ljq−1 of the base of the qth shell triangle by Remark 73.24. Replacing γ by t(γ 0 ), we
are in the case that i = 0 (see the third paragraph of the proof), finishing the proof. ¤
Illustration 73.26. The following picture shows the displacements of the special
marked point of the qth shell triangle in the proof of Proposition 73.25:
6
∗ ∗
∗ ∗ ∗
∗ ∗ ∗ ∗
∗ ∗ ∗ ∗ 3
∗ ∗ ∗ ∗ ∗ ∗
∗ ∗ ∗ ∗ ∗ ∗ ∗
∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗
∗ ∗ 1 ∗ ∗ ∗ ∗ ∗ ∗
∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗
5 ∗ ∗ ∗ 2 ∗ ∗ ∗ ∗ ∗ 4
We start with a cycle γ ∈ Ch(X 2 ) with f (γ) = q, it contains a point somewhere in the
qth shell triangle say, the point in Position 1. Then we modify γ in such a way that f (γ)
is always q and look at what happens with this point. Replacing γ by a maximal order
derivative, we move the special point to the base of the shell triangle; for example, we can
move it to Position 2. The heart of the proof is the movement from Position 2 to Position
3 (here we make use of the induction hypothesis). Again taking an appropriate derivative,
we come to Position 4. Transposing this cycle, we come to Position 5. Finally, repeating
the procedure used in the passage from 2 to 3, we move from Position 5 to Position 6,
arriving to the top.
Illustration 73.27. Let
pr 2
Ch(X 2 ) ,→ Ch(XF2 (X) ) −−−∗→ Ch(X12 )
be the homomorphism used in the proof of Proposition 73.25. If α ∈ Ch(X 2 ), the diagram
of pr 2∗ (α) is obtained from the diagram of α by erasing of the first shell triangle. An
example is shown by the illustration with X as in Remark 73.14:
◦ ◦◦◦◦◦
◦ ◦◦◦◦◦◦◦◦◦◦◦◦◦◦ ◦◦◦◦◦
◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦
◦ ◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦ ◦ ◦◦◦◦◦◦◦◦◦◦◦◦◦◦
◦ ◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦∗◦∗∗∗∗∗ ◦ ◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦∗◦∗∗∗∗∗
∗ ∗∗◦◦◦◦◦◦◦◦◦◦◦◦◦◦∗◦∗◦∗◦◦◦◦◦◦◦◦∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗ ◦ ◦◦◦◦∗◦∗∗◦◦◦◦◦◦◦◦◦∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗
∗ •∗ ∗∗ •∗◦∗ •∗ ◦∗ •∗ ∗∗ •∗ ◦∗ •∗∗∗ •∗ ∗∗ •∗ ∗∗•∗ ∗ •∗ ◦∗•∗ ∗∗ •∗ ◦∗ •∗∗∗ •∗ ∗∗•∗ ∗∗ •∗
diagram of α diagram of pr 2∗ (α)
Summarizing, we have the following structure result on Che(X 2 ):

Theorem 73.28. Let X be a smooth anisotropic quadric. The set of primordial cycles
Π lying in Che(X 2 ) has the following properties:
(1) All derivatives of all cycles in Π are minimal and pairwise disjoint, and the set
of these form a basis of Che(X 2 ). In particular, the sum of all maximal order
derivatives of the elements in Π is equal to the cycle
P
d
∆= (hi × li + li × hi ) ∈ Ch(X̄ 2 ).
i=0
(2) Every cycle in Π is symmetric and has no points outside of the shell triangles.
(3) The map f in Proposition 73.25 is injective on Π. Every cycle π ∈ Π contains
the top of the f (π)th shell triangle and has no points in any shell triangle with
number¡ in f (Π)
¢ \ {f (π)}.
2
(4) 1 ∈ f Ch(X ) = f (Π).
Definition 73.29. Let f be as in Proposition 73.25. If f (α) = q for an element
α ∈ Ch(X 2 ), we say that α starts in the qth shell triangle. More specifically, if f (π) = q
for a primordial cycle π, we say that π is q-primordial.
The following statement is an additional property of 1-primordial cycles:
Proposition 73.30. Let π ∈ Ch(X 2 ) be a 1-primordial cycle. Suppose π contains
hi ×li+i1 −1 for some positive i ≤ d. Then the smallest integer i with this property coincides
with an absolute Witt index of ϕ, i.e., i = jq−1 for some q ∈ [2, h].
Proof. The cycle π contains h0 × li1 −1 as this is the top of the first shell triangle.
By Lemma 73.13, π contains none of cycles h1 × li1 , . . . , hi1 −1 × l2i1 −2 . It follows that if
i ∈ [1, d] is the smallest integer satisfying hi × li+i1 −1 ∈ π then i ≥ j1 = i1 . Let q ∈ [2, h]
be the largest integer with jq−1 ≤ i. We show jq−1 = i. Suppose to the contrary that
jq−1 < i.
Let X1 be the quadric over F (X) given by the anisotropic part of ϕF (X) and
pr 2∗ : Ch(XF2 (X) ) → Ch(X12 )
the homomorphism of Remark 72.5. Then the element pr 2∗ (π) starts in the shell triangle
number q − 1 of X1 . Therefore, by Proposition 73.25, the quadric X1 possesses a (q − 1)-
primordial cycle τ .
Let
in 2∗ : Ch(X12 ) → Ch(XF2 (X) )
be the homomorphism of Remark 72.5. Then the cycle β = in 2∗ (τ ) in Ch(XF2 (X) ) contains
hjq−1 × ljq −1 and does not contain any hj × l? with j < jq−1 .
Let η ∈ Ch(X 3 ) be a preimage of β under the pull-back epimorphism
g ∗ : Ch(X 3 ) ³ Ch(XF2 (X) ) ,
where the morphism g : XF2 (X) → X 3 is induced by the generic point of the first factor of
X 3 (cf. Corollary 57.10). The cycle η contains h0 × hjq−1 × ljq −1 and does not contain any
h0 × hj × l? with j < jq−1 .
We consider η as a correspondence X Ã X 2 . Define µ as the composition µ = η ◦ α
with α = π ·(h0 ×hi1 −1 ). The cycle α contains h0 ×l0 and does not contain any hj ×lj with
j ∈ [1, i). In particular, since jq−1 < i, it does not contain any hj × lj with j ∈ [1, jq−1 ].
Consequently, the cycle µ contains the basis element
h0 × hjq−1 × ljq −1 = (h0 × hjq−1 × ljq −1 ) ◦ (h0 × l0 )
and does not contain any hj × h? × l? with j ∈ [1, jq−1 ].
Let
δ ∗ : Ch(X 3 ) → Ch(X 2 )
be the pull-back homomorphism with respect to the partial diagonal morphism
δ : X 2 → X 3 , (x1 × x2 ) 7→ (x1 × x1 × x2 ) .
The cycle δ ∗ (µ) ∈ Ch(X 2 ), contains the basis element
hjq−1 × ljq −1 = δ ∗ (h0 × hjq−1 × ljq −1 )
and does not contain any hj × l? with j < jq−1 . It follows that an appropriate derivative
of the cycle δ ∗ (µ) contains hi × li+i1 −1 ∈ π and does not contain h0 × li1 −1 ∈ π. This
contradicts the minimality of π. ¤
Remark 73.31. In the language of diagrams, Proposition 73.30 asserts that the point
i
h × li+i1 −1 lies on the left side of the qth shell triangle.
Definition 73.32. We say that the integer q ∈ [2, h] occurring in Proposition 73.30
is produced by the 1-primordial cycle π.
CHAPTER XIV
The Izhboldin dimension
Let X be an anisotropic smooth projective quadric over a field F (of arbitrary char-
acteristic). The Izhboldin dimension dimIzh X of X is defined as
dimIzh X := dim X − i1 (X) + 1 ,
where i1 (X) is the first Witt index of X.
Let Y be a complete (possibly singular) algebraic variety over F with all of its closed
points of even degree and such that Y has a closed point of odd degree over F (X). The
main theorem of this chapter is Theorem 76.1 below. It states that dimIzh X ≤ dim Y
and if dimIzh X = dim Y the quadric X is isotropic over F (Y ).
An application of Theorem 76.1 is the positive solution of a conjecture of Izhboldin that
states: if an anisotropic quadric Y becomes isotropic over F (X) then dimIzh X ≤ dimIzh Y
with equality if and only if X is isotropic over F (Y ).
The results of this chapter for the case that characteristic 6= 2 were obtained in [72].
74. The first Witt index of subforms

For the reader’s convenience, we list some easy properties of the first Witt index:
Lemma 74.1. Let ϕ be an anisotropic non-degenerate quadratic form over F of di-
mension at least two.
(1) The first Witt index i1 (ϕ) coincides with the minimal Witt index of ϕE , where E
runs over all field extension of F such that the form ϕE is isotropic.
(2) Let ψ be a non-degenerate subform of ϕ of codimension r and E/F a field ex-
tension. Then i0 (ψE ) ≥ i0 (ϕE ) − r. In particular, i1 (ψ) ≥ i1 (ϕ) − r (if i1 (ψ) is
defined, i.e., dim ψ ≥ 2).
Proof. The first statement is proven in Corollary 25.3. For the second statement,
note that the intersection of a maximal isotropic subspace U (of dimension i0 (ϕE )) of the
form ϕE with the space of the subform ψE is of codimension at most r in U . ¤
The following two statements are due to Vishik (in the case that characteristic 6= 2,
cf. [140, Cor. 4.9]).
Proposition 74.2. Let ϕ be an anisotropic non-degenerate quadratic form over F
with dim ϕ ≥ 2. Let ψ be a non-degenerate subform of ϕ. If codimϕ ψ ≥ i1 (ϕ) then the
form ψF (ϕ) is anisotropic.
Proof. Let n = codimϕ ψ and assume that n ≥ i1 (ϕ). If the form ψF (ϕ) is isotropic
there exists a rational morphism X 99K Y , where X and Y are the projective quadrics of ϕ
and ψ respectively. We use the same notation as in §72. Let α ∈ Ch(X 2 ) be the class of the
333
334 XIV. THE IZHBOLDIN DIMENSION
closure of the graph of the composition X 99K Y ,→ X. Since the push-forward of α with
respect to the first projection X 2 → X is non-zero, we have h0 ×l0 ∈ α. On the other hand,
as α is in the image of the push-forward homomorphism Ch(X̄ × Ȳ ) → Ch(X̄ 2 ) that maps
any external product β × γ to β × in ∗ (γ), where the push-forward in ∗ : Ch(Ȳ ) → Ch(X̄)
maps hi to hi+n , and n ≥ i1 (ϕ), one has li1 (ϕ)−1 × hi1 (ϕ)−1 ∈
/ α, contradicting Lemma 73.21
(cf. also Corollary 73.23). ¤
Corollary 74.3. Let ϕ be an anisotropic non-degenerate quadratic form and ϕ0 a
non-degenerate subform of ϕ of codimension n with dim ϕ0 ≥ 2. If n < i1 (ϕ) then i1 (ϕ0 ) =
i1 (ϕ) − n.
Proof. Let i1 = i1 (ϕ). By Lemma 74.1, we know that i1 (ϕ0 ) ≥ i1 − n. Let ψ be a
non-degenerate subform of ϕ0 of dimension dim ϕ − i1 . If i1 (ϕ0 ) > i1 − n then the form
ψF (ϕ) is isotropic by Lemma 74.1 contradicting Proposition 74.2. ¤
Lemma 74.4. Let ϕ be an anisotropic non-degenerate quadratic F -form with dim ϕ ≥ 3
and i1 (ϕ) = 1. Let F (t)/F be a purely transcendental field extension of degree 1. Then
there exists a non-degenerate subform ψ of ϕF (t) of codimension one satisfying i1 (ψ) = 1.
Proof. First consider the case that char(F ) 6= 2. We can write ϕ ' ϕ0 ⊥ ha, bi for
some a, b ∈ F × and some quadratic form ϕ0 . Set
®
ψ = ϕ0F (t) ⊥ a + bt2 .
This is clearly a subform of ϕF (t) of codimension 1. Moreover, the fields F (t)(ψ) and F (ϕ)
are isomorphic over F . In particular,
i1 (ψ) = i0 (ψF (t)(ψ) ) ≤ i0 (ϕF (t)(ψ) ) = i0 (ϕF (ϕ) ) = i1 (ϕ) = 1
hence i1 (ψ) = 1.
Suppose char(F ) = 2. If dim ϕ is even then ϕ ' ϕ0 ⊥[a, b] for some a, b ∈ F and some
even-dimensional non-degenerate quadratic form ϕ0 by Proposition 7.32. In this case set
®
ψ = ϕ0F (t) ⊥ a + t + bt2 .
If dim ϕ is odd then ϕ ' ϕ0 ⊥[a, b]⊥ hci for some c ∈ F × , some a, b ∈ F , and some
even-dimensional non-degenerate quadratic form ϕ0 . In this case set
ψ = ϕ0F (t) ⊥[a, b + ct2 ] .
In either case, ψ is a non-degenerate subform of ϕF (t) of codimension 1 such that the
fields F (t)(ψ) and F (ϕ) are F -isomorphic. Therefore the argument above shows that
i1 (ψ) = 1. ¤
75. Correspondences
Let X and Y be schemes over a field F . Suppose that X is equidimensional and let
d = dim X. Recall that a correspondence (of degree zero) α : X Ã Y from X to Y is an
element α ∈ CHd (X × Y ) (cf. §62). A correspondence is called prime if it is represented
by a prime cycle. Every correspondence is a linear combination of prime correspondences
with integer coefficients.
75. CORRESPONDENCES 335
Let α : X Ã Y be a correspondence. Assume that X is a variety and Y is complete.

The projection morphism p : X ×Y → X is proper hence the push-forward homomorphism
p∗ : CHd (X × Y ) → CHd (X) = Z · [X]
is defined (cf. Proposition 49.9). The number mult(α) ∈ Z satisfying p∗ (α) = mult(α)·[X]
is called the multiplicity of α. Clearly, mult(α + β) = mult(α) + mult(β) for any two
correspondences α, β : X Ã Y .
A correspondence α : Spec F Ã Y is represented by a 0-cycle z on Y . Clearly,
mult(α) = deg(z), where deg : CH0 (Y ) → Z is the degree homomorphism defined in
Example 57.6. More generally, we have the following statement.
Lemma 75.1. The composition
deg
CHd (X × Y ) → CH0 (YF (X) ) −−→ Z,
where the first map is the pull-back homomorphism with respect to the natural flat mor-
phism YF (X) → X × Y takes a correspondence α to mult(α).
Proof. The statement follows by Proposition 49.20 applied to the fiber product di-
agram
YF (X) −−−→ X × Y
 
 
y y
Spec F (X) −−−→ X
¤
Lemma 75.2. Let Y be a complete scheme and F̃ /F a purely transcendental field
extension. Then
deg CH0 (Y ) = deg CH0 (YF̃ ) .
Proof. We may assume that F̃ is the function field of the affine line A1 . By Propo-
sition 49.20 applied to the fiber product diagram
f
YF (A1 ) −−−→ Y
 
 
y y
Spec F (A1 ) −−−→ Spec F
it suffices to show the the pull-back homomorphism f ∗ : CH0 (Y ) → CH0 (YF (A1 ) ) is sur-
jective. We have f = p ◦ q where p : Y × A1 → Y is the projection and q : YF (A1 ) → Y × A1
is a natural morphism. The pull-back homomorphism p∗ is an isomorphism by Theorem
57.12 and q ∗ is surjective by Corollary 57.10. As f ∗ = q ∗ ◦ p∗ by Proposition 49.18, the
map f ∗ is surjective. ¤
Corollary 75.3. Let Y be a complete variety, X a smooth projective quadric, and
0
X ⊂ X an arbitrary closed subvariety of X. Then
deg CH0 (YF (X) ) ⊂ deg CH0 (YF (X 0 ) ) .
Proof. Since F (X) is a subfield of F (X × X 0 ), we have

deg CH0 (YF (X) ) ⊂ deg CH0 (YF (X×X 0 ) ) .
As the quadric XF (X 0 ) is isotropic, the field extension F (X × X 0 )/F (X 0 ) is purely tran-
scendental. Hence
deg CH0 (YF (X×X 0 ) ) = deg CH0 (YF (X 0 ) )
by Lemma 75.2. ¤
Let X and Y be varieties over F with dim X = d and Z ⊂ X ×Y a prime d-dimensional
cycle of multiplicity r > 0. The generic point of Z defines a degree r closed point of the
generic fiber YF (X) of the projection X × Y → X and vice versa. Hence there is a natural
bijection of the following two sets for every r > 0:
1) prime d-dimensional cycles on X × Y of multiplicity r.
2) closed points of YF (X) of degree r.
A rational morphism X 99K Y defines a prime correspondence X Ã Y of multiplicity
1 by taking the closure of its graph. Conversely, a prime cycle Z ⊂ X × Y of multiplicity
1 is birationally isomorphic to X. Therefore, the projection to Y defines a rational map
X 99K Z → Y . Hence there are natural bijections between the sets of:
0) rational morphisms X 99K Y .
1) prime d-dimensional cycles on X × Y of multiplicity 1.
2) rational points of YF (X) .
A prime correspondence X Ã Y of multiplicity r can be viewed as a “generically
r-valued map” between X and Y .
Let α : X Ã Y be a correspondence between varieties of dimension d. We write
t
α : Y Ã X for the transpose of α (cf. §62).
Theorem 75.4. Let X be an anisotropic smooth projective quadric with i1 (X) = 1.
Let δ : X Ã X be a correspondence. Then mult(δ) ≡ mult(δ t ) (mod 2).
Proof. The coefficient of h0 × l0 in the decomposition of the class of δ in the modulo
2 reduced Chow group Ch(X 2 ) is mult(δ) (mod 2) (taking into account Lemma 73.2 in
the case when dim X = 0). Therefore, the theorem is a particular case of Corollary 73.23.
(It is also a particular case of Lemma 73.21 and also of Lemma 73.17.)
We give another proof of Theorem 75.4. By Example 66.6, we have
L L
CHd (X 2 ) ' CHd (X) CHd−1 (F l) CH0 (X) ,
where F l is the flag variety of pairs (L, P ), where L and P are a totally isotropic line and
plane, respectively, satisfying L ⊂ P . It suffices to check the formula of Theorem 75.4 for
δ lying in the image of any of these three summands.
Since the embedding CHd (X) ,→ CHd (X 2 ) is given by the push-forward with respect
to the diagonal map, its image is generated by the diagonal class for which the congruence
clearly holds.
Since X is anisotropic, every element of CH0 (X) becomes divisible by 2 over an ex-
tension of F by Theorem 71.2 and Proposition 68.1. As multiplicity is not changed under
a field extension homomorphism, we have mult(δ) ≡ 0 ≡ mult(δ t ) (mod 2) for any δ in
the image of CH0 (X).
76. THE MAIN THEOREM 337
Since the embedding CHd−1 (F l) ,→ CHd (X 2 ) is produced by a correspondence F l Ã

X 2 of degree one, the image of CHd−1 (F l) is contained in the image of the push-forward
CHd (F l × X 2 ) → CHd (X 2 ) with respect to the projection. Let δ ∈ CHd (X 2 ). By Lemma
75.1, the multiplicity of δ and of δ t is the degree of the image of δ under the pull-back
homomorphism CHd (X 2 ) → CH0 (XF (X) ), given by the generic point of the appropriately
chosen factor of X 2 . As i1 (X) = 1, the degree of any closed point on (F l × X)F (X) is even
by Corollary 71.3. Consequently mult(δ) ≡ 0 ≡ mult(δ t ) (mod 2) for any δ in the image
of CHd−1 (F l). ¤
Remark 75.5. The statement of Theorem 75.4 with X replaced by an anisotropic
non smooth quadric (over a field of characteristic 2) was proved by Totaro in [138].
Corollary 75.6. Let X be as in Theorem 75.4. Then any rational endomorphism
f : X 99K X is dominant. In particular, the only point x ∈ X admitting an F -embedding
F (x) ,→ F (X) is the generic point of X.
Proof. Let δ : X Ã X be the class of the closure of the graph of f . Then mult(δ) = 1.
Therefore, the integer mult(δ t ) is odd by Theorem 75.4. In particular, mult(δ t ) 6= 0, i.e.,
f is dominant. ¤
76. The main theorem

The main theorem of the chapter is
Theorem 76.1. Let X be an anisotropic smooth projective quadric over F and Y a
complete variety over F such that every closed point of Y is of even degree. If there is a
closed point in YF (X) of odd degree then
(1) dimIzh X ≤ dim Y .
(2) If dimIzh X = dim Y then X is isotropic over F (Y ).
Proof. A closed point of Y over F (X) of odd degree gives rise to a prime correspon-
dence α : X Ã Y of odd multiplicity. By Springer’s theorem 71.3, to prove statement (2),
it suffices to find a closed point of XF (Y ) of odd degree, equivalently, to find a correspon-
dence Y Ã X of odd multiplicity.
First assume that i1 (X) = 1, so dimIzh X = dim X. In this special case, we simulta-
neously prove both statements of Theorem 76.1 by induction on n = dim X + dim Y .
If n = 0, i.e., X and Y are both of dimension zero then X = Spec K and Y = Spec L
for some field extensions K and L of F with [K : F ] = 2 and [L : F ] even. Taking the
push-forward to Spec(F ) of the correspondence α, we have
[K : F ] · mult(α) = [L : F ] · mult(αt ).
Since mult(α) is odd, the correspondence αt : Y Ã X is of odd multiplicity as needed.
So we may assume that n > 0. Let d be the dimension of X. We first prove (2),
so we have dim Y = d > 0. It suffices to show that mult(αt ) is odd. Assume that the
multiplicity of αt is even. Let x ∈ X be a closed point of degree 2. Since the multiplicity
of the correspondence [Y × x] : Y Ã X is 2 and the multiplicity of [x × Y ] : X Ã Y is zero,
modifying α by adding an appropriate multiple of [x × Y ] we can assume that mult(α) is
odd and mult(αt ) = 0.
The degree of the pull-back of αt on XF (Y ) is now zero by Lemma 75.1. By Corollary

71.4, the degree homomorphism
deg : CH0 (XF (Y ) ) → Z
is injective. Therefore, by Proposition 52.9, there is a nonempty open subset U ⊂ Y such
that the restriction of α on X × U is trivial. Write Y 0 for the reduced scheme Y \ U , and
let i : X × Y 0 → X × Y and j : X × U → X × Y denote the closed and open embeddings
respectively. The sequence
i j∗
CHd (X × Y 0 ) −
→∗
CHd (X × Y ) −
→ CHd (X × U )
is exact by Proposition 57.8. Hence there exists an α0 ∈ CHd (X ×Y 0 ) such that i∗ (α0 ) = α.
We can view α0 as a correspondence X Ã Y 0 . Clearly, mult(α0 ) = mult(α), hence
mult(α0 ) is odd. Since α0 is a linear combination of prime correspondences, there exists
a prime correspondence β : X Ã Y 0 of odd multiplicity. The class β is represented by
a prime cycle, hence we may assume that Y 0 is irreducible. Since dim Y 0 < dim Y =
dim X = dimIzh X, we contradict statement (1) for the varieties X and Y 0 that holds by
the induction hypothesis.
We now prove (1) when i1 (X) = 1. Assume that dim Y < dim X. Let Z ⊂ X × Y be
a prime cycle representing α. Since mult(α) is odd, the projection Z → X is surjective
and the field extension F (X) ,→ F (Z) is of odd degree. The restriction of the projection
X × Y → Y defines a proper morphism Z → Y . Replacing Y by the image of this
morphism, we assume that Z → Y is a surjection.
In view of Lemma 74.4, extending scalars to a purely transcendental extension of F , we
can find a smooth subquadric X 0 of X of the same dimension as Y having i1 (X 0 ) = 1. By
Lemma 75.2, all closed points on Y are still of even degree. Since purely transcendental
extensions do not change Witt indices by Lemma 7.16, we still have i1 (X) = 1.
By Corollary 75.3, there exists a correspondence X 0 Ã Y of odd multiplicity. Since
dim X 0 < dim X, by the induction hypothesis, statement (2) holds for X 0 and Y , that is,
X 0 has a point over Y , i.e., there exists a rational morphism Y 99K X 0 . Composing this
morphism with the embedding of X 0 into X, we get a rational morphism f : Y 99K X.
Consider the rational morphism
h := idX × f : X × Y 99K X × X.
As the projection of Z to Y is surjective, Z intersects the domain of definition of h.
Let Z 0 be the closure of the image of Z under h. The composition of Z 99K Z 0 with
the first projection to X yields a tower of field extensions F (X) ⊂ F (Z 0 ) ⊂ F (Z). As
[F (Z) : F (X)] is odd, so is [F (Z 0 ) : F (X)], i.e., the correspondence β : X Ã X given by
Z 0 is of odd multiplicity. The image of the second projection Z 0 → X is contained in X 0
hence mult(β t ) = 0 as dim X 0 < dim X. This contradicts Theorem 75.4 and establishes
Theorem 76.1 in the case i1 (X) = 1.
We now consider the general case. Let X 0 be a smooth subquadric of X with dim X 0 =
dimIzh X. Then i1 (X 0 ) = 1 by Corollary 74.3, i.e., dimIzh X 0 = dimIzh X. By Corollary
75.3, the scheme YF (X 0 ) has a closed point of odd degree since YF (X) does. As i1 (X 0 ) = 1,
we have shown in the first part of the proof that the statements (1) and (2) hold for X 0
and Y . In particular, dimIzh X = dimIzh X 0 ≤ dim Y by (1) for X 0 and Y proving (1) for
76. THE MAIN THEOREM 339
X and Y . If dim X 0 = dim Y , it follows from (2) applied to X 0 and Y that X 0 is isotropic
over F (Y ). Hence X is isotropic over F (Y ) proving (2) for X and Y . ¤
Remark 76.2. The statement of Theorem 76.1 with X replaced by an anisotropic
non smooth quadric (over a field of characteristic 2) was proved by Totaro in [138].
A consequence of Theorem 76.1 is that an anisotropic smooth quadric X cannot be
compressed to a variety Y of dimension smaller than dimIzh X with all closed points of
even degree:
Corollary 76.3. Let X be an anisotropic smooth projective F -quadric and Y a
complete F -variety with all closed points of even degree. If dimIzh X > dim Y then there
are no rational morphisms X 99K Y .
Remark 76.4. Let X and Y be as in part (2) of Theorem 76.1. Suppose in addition
that dim X = dimIzh X, i.e., i1 (X) = 1. Let α : X Ã Y be a correspondence of odd
multiplicity. The proof of Theorem 76.1 shows mult(αt ) is also odd.
Applying Theorem 76.1 to the special (but perhaps the most interesting) case where
the variety Y is also a projective quadric, we prove the conjectures of Izhboldin:
Theorem 76.5. Let X and Y be anisotropic smooth projective quadrics over F . Sup-
pose that Y is isotropic over F (X). Then
(1) dimIzh X ≤ dimIzh Y .
(2) dimIzh X = dimIzh Y if and only if X is isotropic over F (Y ).
Proof. Choose a subquadric Y 0 ⊂ Y with dim Y 0 = dimIzh Y . Since Y 0 becomes
isotropic over F (Y ) by Lemma 74.1(2) and Y becomes isotropic over F (X), the quadric
Y 0 becomes isotropic over F (X). By Theorem 76.1, we have dimIzh X ≤ dim Y 0 . Moreover,
in the case of equality, X becomes isotropic over F (Y 0 ) and hence over F (Y ). Conversely,
if X is isotropic over F (Y ), interchanging the roles of X and Y , the argument above also
yields dimIzh Y ≤ dimIzh X, hence equality holds. ¤
We have the following upper bound for the Witt index of Y over F (X).
Corollary 76.6. Let X and Y be anisotropic smooth projective quadrics over F .
Suppose that Y is isotropic over F (X). Then
i0 (YF (X) ) − i1 (Y ) ≤ dimIzh Y − dimIzh X .
Proof. If dimIzh X = 0, the statement is trivial. Otherwise, let Y 0 be a smooth
subquadric of Y of dimension dimIzh X − 1. Since dimIzh Y 0 ≤ dim Y 0 < dimIzh X, the
quadric Y 0 remains anisotropic over F (X) by Theorem 76.5(1). Therefore, i0 (YF (X) ) ≤
codimY Y 0 = dim Y − dimIzh X + 1 by Lemma 74.1, hence the inequality. ¤
We have also the following more precise version of Theorem 76.1:
Corollary 76.7. Let X be an anisotropic smooth projective F -quadric and Y a
complete variety over F such that every closed point of Y is of even degree. If there is a
closed point in YF (X) of odd degree then there exists a closed subvariety Y 0 ⊂ Y such that
(i) dim Y 0 = dimIzh X.
(ii) YF0 (X) possesses a closed point of odd degree.

(iii) XF (Y 0 ) is isotropic.
Proof. Let X 0 ⊂ X be a smooth subquadric with dim X 0 = dimIzh X. Then
dimIzh X 0 = dim X 0 by Corollary 74.3. An odd degree closed point on YF (X) determines a
correspondence X Ã Y of odd multiplicity which in turn gives a correspondence X 0 Ã Y
of odd multiplicity. We may assume that the latter correspondence is prime and take a
prime cycle Z ⊂ X 0 × Y representing it. Let Y 0 be the image of the proper morphism
Z → Y . Clearly, dim Y 0 ≤ dim Z = dim X 0 = dimIzh X. On the other hand, Z determines
a correspondence X 0 Ã Y 0 of odd multiplicity. Therefore dim Y 0 ≥ dim X 0 by Theorem
76.1, and condition (i) of Corollary 76.7 is satisfied. Moreover, YF0 (X 0 ) has a closed point
of odd degree. Since the field F (X × X 0 ) is a purely transcendental extension of F (X),
Lemma 75.2 shows that there is a closed point on YF0 (X) of odd degree, i.e., condition (ii)
of Corollary 76.7 is satisfied. Finally, the quadric XF0 (Y 0 ) is isotropic by Theorem 76.1;
therefore XF (Y 0 ) is isotropic. ¤
77. Addendum: The Pythagoras number
Given a field F , its pythagoras number is defined to be

p(F ) := min{n | D(nh1i) = D(∞h1i)}
or infinity if no such integer exists. If char F = 2 then p(F ) = 1 and if char F 6= 2 then
p(F ) = 1 if and only if F is pythagorean. Let F be a field that is not formally real. Then
the quadratic form (s(F ) + 1) h1i is isotropic. In particular, p(F ) = s(F ) or s(F ) + 1 and
each value is possible. So this invariant is only interesting when the field is formally real.
For a given formally real field determining its pythagoras number is not easy. If F is an
extension of a real closed field of transcendence degree n then p(F ) ≤ 2n by Corollary
35.15.
¡ In particular,
¢ if n = 1 and F is not pythagorean then¡ p(F ) = 2. ¢ It is known that
p R(t1 , t2 ) = 4 (cf. [25]), but in general, the value of p R(t1 , . . . tn ) is not known. In
this section, given any non-negative integer n, we construct a formally real field having
pythagoras number n. This was first done by Hoffmann in [57].
Lemma 77.1. Let F be a formally real field and ϕ a quadratic form over F . If P ∈
X(F ) then P extends to an ordering on F (ϕ) if and only if ϕ is indefinite at P , i.e.,
|sgnP (ϕ)| < dim ϕ.
Proof. Suppose that ϕ is indefinite at P . Let FP be the real closure of F with respect
to P . Let K = FP (ϕ). As ϕFP is isotropic, K/FP is purely transcendental. Therefore the
unique ordering on FP extends to K. The restriction of this extension to F (ϕ) extends
P . The converse is clear. ¤
The following proposition is a consequence of the lemma and Theorem 76.5.
Proposition 77.2. Let F be formally real and x, y ∈ D(∞h1i). Let ϕ ' mh1i ⊥ h−xi
and ψ ' nh1i ⊥ h−yi with n > m ≥ 0. Then F (ψ) is formally real. If, in addition, ϕ is
anisotropic then so is ϕF (ψ) .
77. ADDENDUM: THE PYTHAGORAS NUMBER 341
Proof. As ψ is indefinite at every ordering, every ordering of F extends to F (ψ).

In particular, F (ψ) is formally real. Suppose that ϕ is anisotropic. Since over each real
closure of F both ϕ and ψ have Witt index 1, the first Witt index of ϕ and ψ must also
be one. As dim ϕ > dim ψ, the form ϕF (ψ) is anisotropic by Theorem 76.5. ¤
Construction 77.3. Let F0 be a formally real field. Let F1 = F0 (t1 , . . . , tn−1 ) and
x = 1 + t21 + · · · + t2n−1 ∈ D(∞h1i). By Corollary 17.13, the element x is a sum of n
squares in F1 but no fewer. In particular, ϕ ' (n − 1)h1i ⊥ h−xi is anisotropic over F1 .
For i ≥ 1, inductively define Fi+1 as follows:
Let
Ai := {nh1i ⊥ h−yi | y ∈ D(∞h1iFi ) }.
For any finite subset S ⊂ Ai , let XS be the product of quadrics Xϕ for all ϕ ∈ S. If
S ⊂ T are two subsets of Ai , we have the dominant projection XT → XS and therefore
the inclusion of function fields F (XS ) → F (XT ). Set Fi+1 = colim FS over all finite
subsets S ⊂ Ai . By construction,
S all quadratic forms ϕ ∈ Ai are isotropic over the field
extension Fi of F . Let F = Fi . Then F has the following properties.
(1) F is formally real. P
(2) nh1i ⊥ h−yi is isotropic for all 0 6= y ∈ (F × )2 .
Consequently, D(∞h1iF ) ⊂ D(nh1iF ), so the pythagoras number p(F ) ≤ n. As ϕ '
(n − 1) h1i ⊥ h−xi remains anisotropic over F , we have p(F ) ≥ n. So we have shown
Theorem 77.4. For every n ≥ 1 there exists a formally real field F with p(F ) = n.
CHAPTER XV
Application of Steenrod operations
Since Steenrod operations are not available in characteristic 2, throughout this chapter,
the characteristic of the base field is assumed to be different from 2.
We write v2 (n) for the 2-adic exponent of an integer n.
We shall use the notation of Chapter XIII. In particular, X is a smooth D-dimensional
projective quadric over a field F given by a (non-degenerate) quadratic form ϕ, and
d = [D/2].
78. Computation of Steenrod operations

Recall that h ∈ Ch1 (X) is the class of a hyperplane section.
Lemma 78.1. The modulo 2 total Chern class c(TX ) : Ch(X) → Ch(X) of the tangent
vector bundle TX of the quadric X is multiplication by (1 + h)D+2 .
Proof. By Proposition 58.15, it suffices to show that c(TX )([X]) = (1 + h)D+2 .
Let i : X ,→ P be the closed embedding of X into the (D + 1)-dimensional projective
space P = P(V ), where V is the underlying vector space of ϕ. We write H ∈ Ch1 (P)
for the class of a hyperplane, so h = i∗ (H). Since X is a hypersurface in P of degree
2, the normal bundle N of the embedding i is isomorphic to i∗ (L⊗2 ), where L is the
canonical¡ line bundle
¢ over P (cf. §104.B). By Propositions 104.16 and 54.7, we have
c(TX ) ◦ c i (L ) = c(i∗ TP ). By Example 61.15, we know that c(TP ) is the multiplication
∗ ⊗2
by (1 + H)D+2 and by Propositions 54.3 and 57.25 that c(L⊗2 ) = id modulo 2. It follows
that
¡ ¡ ¢−1 ¢ ∗
c(TX )([X]) = c(i∗ TP ) ◦ c i∗ (L⊗2 ) (i ([P])) =
¡∗ ¢
i ◦ c(TP ) ◦ c(L⊗2 )−1 ([P]) = i∗ (1 + H)D+2 = (1 + h)D+2
by Proposition 55.22. ¤
Corollary 78.2. Suppose that i0 (X) > n for some n ≥ 0. Let W ⊂ V be a totally
isotropic (n + 1)-dimensional subspace of V and P be the n-dimensional projective space
P(W ). Let i : P ,→ X be the closed embedding. Then the modulo 2 total Chern class
c(N ) : Ch(P) → Ch(P) of the normal bundle N of the embedding i is multiplication by
(1 + H)D+1−n , where H ∈ Ch1 (P) is the class of a hyperplane.
Proof. By Propositions 104.16 and 54.7, we have c(N ) = c(TP )−1 ◦ c(i∗ TX ); and by
Proposition 55.22 and Lemma 78.1, we have
¡ ¢
c(i∗ TX )[P] = c(i∗ TX )(i∗ [X]) = i∗ ◦ c(TX ) [X] = i∗ (1 + h)D+2 = (1 + H)D+2 .
Hence by Example 61.15, c(TP ) = (1 + H)n+1 . ¤
343
344 XV. APPLICATION OF STEENROD OPERATIONS
Corollary 78.3. Under the hypothesis of Corollary 78.2, we have

SqX ([P]) = [P] · (1 + h)D+1−n .
¡ ¢
Proof. By the Wu Formula 61.7, we have SqX ([P]) = i∗ c(N )[P] . Using Corollary
78.2, we get
¡ ¢ ¡ ¢ ¡ ¢
i∗ c(N )[P] = i∗ (1 + H)D+1−n · [P] = i∗ i∗ (1 + h)D+1−n · [P] = (1 + h)D+1−n · i∗ [P]
using the Projection Formula 56.9. ¤
We also have (cf. Example 61.15):
Lemma 78.4. SqX (hi ) = hi · (1 + h)i for any i ≥ 0.
Corollary 78.5. If the quadric X is split then the ring endomorphism SqX : Ch(X) →
Ch(X) acts on the basis {hi , li }i∈[0, d] of Ch(X) by the formula
SqX (hi ) = hi · (1 + h)i and SqX (li ) = li · (1 + h)D+1−i .
In particular, for any j ≥ 0
µ ¶ µ ¶
j i i+j D+1−i
i
SqX (h ) = h and SqjX (li ) = li−j .
j j
Binomial coefficients modulo 2 are computed as follows (we leave the proof to the
reader): Let N be the set of non-negative integers, 2N the set of all subsets of N, and let
π : N → 2N be the injection given by base 2 expansions. For any n ∈ N, the set π(n)
consists of all those m ∈ N such that the base 2 expansion of n has 1 on the mth position.
For two arbitrary non-negative integers i and n, write i ⊂ n if π(i) ⊂ π(n).
¡ ¢
Lemma 78.6. For any i, n ∈ N, the binomial coefficient ni is odd if and only if i ⊂ n.
79. Values of the first Witt index

The main result of this section is Theorem 79.9 (conjectured by Hoffmann and orig-
inally proved in [73]); its main ingredient is given by Proposition 79.4. We begin with
some observations.
Remark 79.1. By Theorem 61.8,
Sq ∗
Ch(X ∗ ) −−−
X
→ Ch(X ∗ )
 
 
y y
Sq ∗
Ch(X̄ ∗ ) −−−→ Ch(X̄ ∗ )
X̄
is commutative, hence we get an endomorphism Ch(X ∗ ) → Ch(X ∗ ) that we shall also

call a Steenrod operation and denote it by SqX ∗ , even though it is a restriction of SqX̄ ∗
and not of SqX ∗ .
Remark 79.2. Let ln × hm ∈ Ch(X̄ 2 ) be an essential basis element with n ≥ m. Since
Sq(ln × hm ) = Sq(ln ) × Sq(hm ) by Theorem 61.13, we see by Corollary 78.5, that the value
of Sq(ln × hm ) is a linear combination of the elements li × hj with i ≤ n and j ≥ m. If
79. VALUES OF THE FIRST WITT INDEX 345
m = 0, one can say more: Sq(ln × h0 ) is a linear combination of the elements li × h0 with
i ≤ n.
Of course, we have similar facts for the essential basis elements of type hm × ln .
Representing essential basis elements of type ln × hm with n ≥ m as points of the right
pyramid of Remark 73.9, we may interpret the above statements graphically as follows:
the diagram of the value of the Steenrod operation on a point ln × hm is contained in the
isosceles triangle based on the lower row of the pyramid whose top is the point ln × hm
(an example of this is the picture on the left below). If ln × hm is on the right side of
the pyramid, then the diagram of the value of the Steenrod operation is contained in the
part of the right side of the pyramid, which is below the point (an example of this is the
picture on the right below).
◦ ◦◦◦◦◦ ◦ ◦◦◦◦◦
◦ ◦◦◦◦◦◦◦◦◦◦◦◦◦◦ ◦ ◦◦◦◦◦◦◦◦◦◦◦◦◦◦
◦ ◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦•◦◦◦◦◦◦ ◦ ◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦•
◦ ◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦∗◦∗∗∗∗∗∗∗∗◦◦◦◦◦◦ ◦ ◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦◦∗◦∗∗
◦ ◦◦ ◦◦ ◦◦◦◦ ◦◦ ◦∗ ∗∗ ∗∗∗∗ ∗∗ ∗∗◦◦ ◦◦ ◦ ◦◦ ◦◦ ◦◦◦◦ ◦◦ ◦◦◦◦ ◦◦ ◦◦ ◦◦ ◦◦◦◦ ∗∗
The next statement follows simply from Remark 79.2.
Lemma 79.3. Assume that X is anisotropic. Let π ∈ ChD+i1 −1 (X 2 ) be the 1-primordial
j
cycle. For any j ≥ 1, the element SX 2 (π) has no points in the first shell triangle.
Proof. By the definition of π, the only point the cycle π has in the first (left as well
as right) shell triangle is the top of the triangle. By Remark 79.2, the only point in the
first left shell triangle that may be contained in S j (π) is the point on the left side of the
triangle and the only point in the first right shell triangle that may be contained in S j (π)
is the point on the right side of the triangle. Since these two points are not dual (points
of the left side of the first left shell triangle are dual to points on the left side of the first
right shell triangle), the lemma follows by Corollary 73.23. ¤
We shall obtain further information in Lemma 83.1 below. We write exp2 (a) for 2a .
Proposition 79.4. For any anisotropic quadratic form ϕ of dim ϕ ≥ 2,
¡ ¢
i1 (ϕ) ≤ exp2 v2 dim ϕ − i1 (ϕ) .
r
Proof. Let r = v2 (dim ϕ − i1 (ϕ)). Apply the Steenrod operation Sq2X 2 : Ch(X 2 ) →
Ch(X 2 ) to the 1-primordial cycle π. Since
µ ¶
2r 0 0 2r dim ϕ − i1
SqX 2 (h × li1 −1 ) = h × SqX (li1 −1 ) = · (h0 × li1 −1−2r )
2r
by Theorem 61.13 and Corollary 78.5 and the binomial coefficient is odd by Lemma 78.6,
r
we have h0 × li1 −1−2r ∈ Sq2X 2 (α). It follows by Lemma 79.3 that 2r 6∈ (0, i1 ), i.e., that
2r ≥ i1 . ¤
Remark 79.5. Let a be a positive integer written in base 2. A suffix of a is an integer
written in base 2 that is obtained from a by deleting several (and at least one) consecutive
digits starting from the left one. For example, all suffixes of 1011010 are 11010, 1010, 10
and 0.
Let i < n be two non-negative integers. Then the following are equivalent.
(1) i ≤ exp2 v2 (n − i).
(2) There exists an r ≥ 0 satisfying 2r < n, i ≡ n (mod 2r ), and i ∈ [1, 2r ].
(3) i − 1 is the remainder upon division of n − 1 by an appropriate 2-power.
(4) The 2-adic expansion of i − 1 is a suffix of the 2-adic expansion of n − 1.
(5) The 2-adic expansion of i is a suffix of the 2-adic expansion of n or i is a 2-power
divisor of n.
In particular, the integers i = i1 (ϕ) and n = dim ϕ in Proposition 79.4 satisfy these
conditions.
Corollary 79.6. All relative higher Witt indices of an odd-dimensional quadratic
form are odd. The relative higher Witt indices of an even-dimensional quadratic form are
either even or one.
Example 79.7. Assume that ϕ is anisotropic and let s ≥ 0 be the largest integer
satisfying dim ϕ > 2s . Then it follows by Proposition 79.4 that i1 (ϕ) ≤ dim ϕ − 2s (e.g.,
by Condition (4) of Remark 79.5). In particular, if dim ϕ = 2s + 1 then i1 (ϕ) = 1.
The first statement of the following corollary is the Separation Theorem 26.5 (over a
field of characteristic not two); the second statement was originally proved by O. Izhboldin
(using a different method) in [65, Th. 02]. A characteristic two version has been proven
by Hoffmann and Laghribi in [60, Th. 1.3].
Corollary 79.8. Let ϕ and ψ be two anisotropic quadratic forms over F .
(1) If dim ψ ≤ 2s < dim ϕ for some s ≥ 0 then the form ψF (ϕ) is anisotropic.
(2) Suppose that dim ψ = 2s + 1 ≤ dim ϕ for some s ≥ 0. If the form ψF (ϕ) is
isotropic then the form ϕF (ψ) is also isotropic.
Proof. Let X and Y be the quadrics of ϕ and of ψ respectively. Then dimIzh X ≥
2s − 1 by Example 79.7. If dim ψ ≤ 2s then dim Y ≤ 2s − 2. Therefore,
dimIzh Y ≤ dim Y < 2s − 1 ≤ dimIzh X
and YF (X) is anisotropic by Theorem 76.5(1).
Suppose that dim ψ = 2s + 1. Then dimIzh Y = 2s − 1 ≤ dimIzh X. If YF (X) is isotropic
then dimIzh Y = dimIzh X by Theorem 76.5(1) and therefore XF (Y ) is isotropic by Theorem
76.5(2). ¤
We show next that all values of the first Witt index not forbidden by Proposition 79.4
are possible by establishing the main result of this section:
Theorem 79.9. Two non-negative integers i and n satisfy i ≤ exp2 v2 (n − i) if and
only if there exists an anisotropic quadratic form ϕ over a field of characteristic not two
with
dim ϕ = n and i1 (ϕ) = i .
80. ROST CORRESPONDENCES 347
Proof. By Proposition 79.4 and Remark 79.5, we need only prove the necessity. Let
i and n be two non-negative integers satisfying i ≤ exp2 v2 (n − i). Let r be as in condition
(2) of Remark 79.5. Write n − i = 2r · m for some integer m.
Let F be any field of characteristic not two and consider the field K = F (t1 , . . . , tr )
of rational functions in r variables. By Corollary 19.6, the Pfister form π = hht1 , . . . , tr ii
is anisotropic over K. Let L = K(s1 , . . . , sm ), where s1 , . . . , sm are variables. By Lemma
19.5, the quadratic L-form ψ = πF ⊗ h1, s1 , . . . , sm i is anisotropic.
We claim that i1 (ψ) = 2r . Indeed,
√ by Proposition 6.22, we have i1 (ψ) ≥ 2r . On
the other hand, the field E = L( −s1 ) is purely transcendental over K(s2 , . . . , sm ) and
therefore i0 (ψE ) = 2r . Consequently, i1 (ψ) = 2r .
Let ϕ be an arbitrary subform of ψ of codimension 2r − i. As dim ψ = 2r · (m + 1) =
n + (2r − i), the dimension of ϕ is equal to n. Since 2r − i < 2r = i1 (ψ), we have i1 (ϕ) = i
by Corollary 74.3. ¤
80. Rost correspondences

Recall that by abuse of notation we also denote the image of the element h ∈ CH1 (X)
in the groups CH1 (X̄), Ch1 (X), and Ch1 (X̄) by the same symbol h. In the following
lemma, h stands for the element of Ch(X).
Lemma 80.1. Let n be the integer satisfying
2n − 1 ≤ D ≤ 2n+1 − 2 .
Set s = D − 2n + 1 and r = 2n+1 − 2 − D (observe that r + s = 2n − 1). If α ∈ Chr+s (X)
then
SqX r 2
r+s (α) = h · α ∈ Ch0 (X) .
Proof. By the definition of the cohomological Steenrod operation SqX , we have

SqX = c(TX ) ◦ SqX , where SqX is the homological Steenrod operation. Therefore,
SqX = c(−TX ) ◦ SqX . In particular,
P
r+s
SqX
r+s (α) = ci (−TX ) ◦ Sqr+s−i
X (α)
i=0
¡ ¢ i ¡−D−2¢ ¡D+i+1¢
in Ch0 (X). By Lemma 78.1, we have ci (−TX ) = −D−2 i
· h . As i
= ± i
, it
follows from Lemma 78.6, that the latter binomial coefficient is even for any i ∈ [r+1, r+s]
and is odd for i = r. Since Sqr+s−i
X (α) is equal to 0 for i ∈ [0, r − 1] and is equal to α2
for i = r by Theorem 61.12, the required relation is established. ¤
Theorem 80.2. Let n be the integer satisfying
2n − 1 ≤ D ≤ 2n+1 − 2 .
Set s = D−2n +1 and r = 2n+1 −2−D. Let X and Y be two anisotropic projective quadrics
of dimension D over a field of characteristic not two. Let ρ̄ ∈ Chr+s (X × Y ). Then
(pr X )∗ (ρ̄) = 0 if and only if (pr Y )∗ (ρ̄) = 0, where pr X : X ×Y → X and pr Y : X ×Y → Y
are the projections.
Proof. Let ρ be an element of the non-reduced Chow group Chr+s (X × Y ). Write ρ̄

for the image of ρ in Chr+s (X ×Y ). The group Chr+s (X) is generated by hs = hsX (if s = d,
this is true as X is anisotropic hence not split). Therefore, we have (pr X )∗ (ρ̄) = aX hsX for
some aX ∈ Z/2Z. Similarly, (pr Y )∗ (ρ̄) = aY hsY for some aY ∈ Z/2Z. To prove Theorem
80.2 we must show aX = aY .
As X and Y are anisotropic, any closed point z in X or Y has even degree. In
particular, the map
1
degZ : Ch0 (Z) → Z/2Z
2
given by z 7→ 12 deg(z) (mod 2) is well defined for Z = X or Y . Consider the following
diagram:
Chr+s (X × Y )
OOO
(pr X )∗ ooo OOO(pr Y )∗
ooooo OOO
ow o O'
Chr+s (X) SqX×Y
r+s Chr+s (Y )
²
(80.3) SqX
r+s Ch0 (X × Y ) SqY
r+s
OOO
(pr X )∗ ooo OOO(pr Y )∗
ooooo OOO
² woo O' ²
Ch0 (X) Ch0 (Y )
OOO 1 deg
OOO2 1
degY ooo
ooo
X 2
OOO oo
OO' ow oo
Z/2Z
We show that the diagram (80.3) is commutative. The bottom diamond is commutative
by the functorial property of the push-forward homomorphism (cf. Proposition 49.9 and
Example 57.6). The left and the right parallelograms are commutative by Theorem 60.5.
Therefore,
( 12 degX ) ◦ SqX 1 Y
r+s ◦(pr X )∗ (ρ) = ( 2 degY ) ◦ Sqr+s ◦(pr Y )∗ (ρ) .
Applying Lemma 80.1 to the element α = (pr X )∗ (ρ), we have
( 21 degX ) ◦ SqX 1 r 2
r+s ◦(pr X )∗ (ρ) = ( 2 degX )(hX · α ) = aX .
Similarly ( 12 degY ) ◦ SqYr+s ◦(pr Y )∗ (ρ) = aY , proving the theorem. ¤

Exercise 80.4. Use Theorem 80.2 to prove the following generalization of Corollary
79.8(2): Let X and Y be two anisotropic projective quadrics satisfying dim X = dim Y =
D. Let s be as in Theorem 80.2. If there exists a rational morphism X 99K Y , then there
exists a rational morphism Gs (Y ) 99K Gs (X) where for an integer i, Gi (X) is the scheme
(variety, if i 6= D/2) of i-dimensional linear subspaces lying on X. (We shall study the
scheme Gd (X) in Chapter XVI.)
Remark 80.5. One can generalize Theorem 80.2 as follows. We replace Y by an
arbitrary projective variety of an arbitrary dimension (and, in fact, Y need not be smooth
nor of dimension D = dim X). Suppose that every closed point of Y has even degree. Let
80. ROST CORRESPONDENCES 349
ρ ∈ Chr+s (X × Y ) satisfy (pr X )∗ (ρ̄) 6= 0 ∈ Ch(X). Then (pr Y )∗ (ρ) 6= 0 ∈ Ch(Y ) (note
that this is in Ch(Y ), not Ch(Y )). To prove this generalization, we use the commutative
diagram 80.3. As before, we have degX ◦ SqX r+s ◦(pr X )∗ (ρ) 6= 0 provided that pr X (ρ̄) 6= 0.
Y
Therefore, degY ◦ Sqr+s ◦(pr Y )∗ (ρ) 6= 0. In particular, (pr Y )∗ (ρ) 6= 0.
Exercise 80.6. Show that one cannot replace the conclusion (pr Y )∗ (ρ) 6= 0 ∈ Ch(Y )
by (pr Y )∗ (ρ̄) 6= 0 ∈ Ch(Y ) in Remark 80.5. (Hint: Let Y be an anisotropic quadric with
X a subquadric of Y satisfying 2 dim X < dim Y and ρ ∈ Ch(X × Y ) the class of the
diagonal of X.)
Taking Y = X in Theorem 80.2, we have
Corollary 80.7. Let X be an anisotropic quadric of dimension D and s as in The-
orem 80.2. If a rational cycle in Ch(X̄ 2 ) contains hs × l0 then it also contains l0 × hs .
Corollary 80.8. Assume that X is an anisotropic quadric of dimension D and for
some integer i ∈ [0, d] the cycle h0 × li + li × h0 ∈ Ch(X̄ 2 ) is rational. Then the integer
dim X − i + 1 is a power of 2.
Proof. If the cycle h0 × li + li × h0 is rational, then, multiplying by hs × hi , we see
that the cycle hs × l0 + li−s × hi is also rational. By Corollary 80.7, it follows that i = s.
Therefore, dim X − i + 1 = 2n with n as in Theorem 80.2. ¤
Remark 80.9. By Lemma 73.13 and Corollary 73.23, the integer i in Corollary 80.8
is necessarily equal to i1 (X) − 1.
Recalling Definition 73.32, we have
Corollary 80.10. If the integer dim ϕ − i1 (ϕ) is not a 2-power then the 1-primordial
cycle on X 2 produces an integer.
Proof. If the 1-primordial cycle π does not produce any
¡ integer¢ then π = h0 × li1 −1 +
li1 −1 × h0 . Therefore, by Corollary 80.8, the integer D − i1 (ϕ) − 1 + 1 = dim ϕ − i1 (ϕ)
is a 2-power. ¤
Definition 80.11. The element h0 × l0 + l0 × h0 ∈ Ch(X̄ 2 ) is called the Rost corre-
spondence of the quadric X.
Of course, the Rost correspondence of an isotropic quadric X is rational. A special
case of Corollary 80.8 is given by:
Corollary 80.12. If X is anisotropic and the Rost correspondence of X is rational
then D + 1 is a power of 2.
By multiplying by h1 × h0 , we see that rationality of the Rost correspondence implies
rationality of the element h1 × l0 . In fact, rationality of h1 × l0 alone implies that D + 1
is a power of 2:
Corollary 80.13. If X is anisotropic and the element h1 × l0 ∈ Ch(X̄ 2 ) is rational
then D + 1 is a power of 2.
Proof. If h1 × l0 is rational, then for any i ≥ 1, the element hi × l0 is also rational.
Let s be as in Theorem 80.2. By Corollary 80.7 it follows that s = 0, i.e., D = 2n − 1. ¤
Let A be a point of a shell triangle of a quadric. We write A] for the dual point in the
sense of Definition 73.22. The following statement was originally proved (in characteristic
0) by Vishik.
Corollary 80.14. Let Y be another anisotropic projective quadric of dimension D
over some field of characteristic not two. Basis elements of Ch(Ȳ 2 ) are in natural 1-1
correspondence with basis elements of Ch(X̄ 2 ). Assume that i1 (Y ) = s + 1 with s is as in
Corollary 80.7. Let A be a point of a first shell triangle of Y and let A] be its dual point.
Then in the diagram of any element of Ch(X 2 ), the point corresponding to A is marked
if and only if the point corresponding to A] is marked.
Proof. We may assume that A lies in the left first shell triangle of Y . Let hi × lj
(with 0 ≤ i ≤ j ≤ s) be the basis element represented by A. Then the basis element
represented by A] is ls−i × hs−j . Let α ∈ Ch(X 2 ) and assume that α contains hi × lj .
Then the rational cycle (hs−i × hj ) · α contains hs × l0 . Therefore, by Corollary 80.7, this
rational cycle also contains l0 × hs . It follows that α contains ls−i × hs−j . ¤
Remark 80.15. The equality i1 (Y ) = s + 1 holds if Y is excellent. By Theorem 79.9
this value of the first Witt index is maximal for all D-dimensional anisotropic quadrics.
81. On the 2-adic order of higher Witt indices, I
The main result of this section is Theorem 81.3 on a relationship between higher Witt
indices and the integer produced by a 1-primordial cycle. This is used to establish a
relationship between higher Witt indices of an anisotropic quadratic form (cf. Corollary
81.20).
Let ϕ be a non-degenerate (possibly isotropic) quadratic form of dimension D over a
field F of characteristic not two and X = Xϕ . Let h = h(ϕ) be the height of ϕ (or X)
and
F = F0 ⊂ F1 ⊂ · · · ⊂ Fh
the generic splitting tower (cf. §25). For q ∈ [0, h], let iq = iq (ϕ), jq = jq (ϕ), ϕq = (ϕFq )an
and Xq = Xϕq .
We shall use the following simple observation in the proof of Theorem 81.3:
Proposition 81.1. Let α be a homogeneous element of Ch(X̄ 2 ) with codim α > d.
Assume X is not split, i.e., h > 0 and that for some q ∈ [0, h − 1] the cycle α is Fq -
rational and does not contain hi × l? or l? × hi for any i < jq . Then δX
∗
(α) = 0 in Ch(X̄),
2
where δX : X → X is the diagonal morphism of X.
∗
Proof. We may assume that dim α = D + i with i ≥ 0 (otherwise dim δX (α) < 0).
As X is not hyperbolic, ld × ld ∈ / α by Lemma 73.2. Therefore, by Proposition 68.1,
∗
δX (α) = nli , where n is the number of essential basis elements contained in α. Since α
does not contain any hi × l? or l? × hi with i < jq , the number of essential basis elements
contained in α coincides with the number of essential basis elements contained in pr 2∗ (α),
where
pr 2∗ : Ch(XF2q ) → Ch(Xq2 )
is the homomorphism of Remark 72.5. In view of Corollary 72.4, the latter number is
even by Lemma 73.16. ¤
81. ON THE 2-ADIC ORDER OF HIGHER WITT INDICES, I 351
We have defined minimal and primordial elements in Ch(X 2 ) for an anisotropic quadric
X (cf. Definitions 73.5 and 73.18). We extend these definitions to the case of an arbitrary
quadric.
Definition 81.2. Let X be an arbitrary (smooth) quadric given by a quadratic form
ϕ (not necessarily anisotropic) and let X0 be the quadric given by the anisotropic part
of ϕ. The images of minimal (respectively primordial) elements via the embedding in 2∗ :
Ch(X02 ) → Ch(X 2 ) of Remark 72.5 are called minimal (respectively primordial) elements
of Ch(X 2 ).
Theorem 81.3. Let X be an anisotropic quadric of even dimension over a field of
characteristic not two. Let π ∈ Ch(X 2 ) be the 1-primordial cycle. Suppose that π produces
an integer q ∈ [2, h] satisfying v2 (i2 + · · · + iq−1 ) ≥ v2 (i1 ) + 2. Then v2 (iq ) ≤ v2 (i1 ) + 1.
Proof. We fix the following notation:
a = i1 ,
b = i2 + · · · + iq−1 = jq−1 − a ,
c = iq .
Set n = v2 (i1 ). So v2 (b) ≥ n + 2.
Consider the (a − 1)th derivative α = π · (h0 × ha−1 ) of π. By Lemma 73.11 and
Proposition 73.30, the cycle α is minimal since π is and contains the basis elements h0 × l0
and ha+b × la+b .
Suppose the result is false, i.e., v2 (c) ≥ n + 2. Proposition 81.4 below contradicts the
minimality of α, hence will prove Theorem 81.3. To state Proposition 81.4, we need the
following morphisms:
g1 : XF2 (X) → X 3
the morphism given by the generic point of the first factor of X 3 ;
t12 : Ch(X 3 ) → Ch(X 3 )
the automorphism given by the transposition of the first two factors of X 3 ;
δX 2 : X 2 → X 4 , (x1 , x2 ) 7→ (x1 , x2 , x1 , x2 )
the diagonal morphism of X 2 . We also use the pairing
◦ : Ch(X̄ r ) × Ch(X̄ s ) → Ch(X̄ r+s−2 )
(for various r, s ≥ 1) given by composition of correspondences, where the elements of
Ch(X̄ s ) are considered as correspondences X̄ s−1 Ã X̄ and the elements of Ch(X̄ r ) are
considered as correspondences X̄ Ã X̄ r−1 .
Note that applying Proposition 73.25 to the quadric X1 with cycle pr 2∗ (π) ∈ Ch(X12 ),
there exists a homogeneous essential symmetric cycle β ∈ Ch(XF2 (X) ) containing the basis
element ha+b × la+b+c−1 and none of the basis elements having hi with i < a + b as a factor.
Proposition 81.4. In the notation above let η ∈ Ch(X 3 ) be a preimage of β under
the pull-back epimorphism g1∗ and µ the essence of the composition η ◦ α. Then the cycle
³ ¡ 2a ¢´
(h0 × hc−a−1 ) · δX
∗
2 t12 (µ) ◦ SqX 3 (µ) · (h0 × h0 × hc−a−1 ) ∈ Ch(X 2 )
contains ha+b × la+b and does not contain h0 × l0 .

Proof. Recall that b ≥ 0 and 2n+2 divides b and c, where n = v2 (a). By Proposition
79.4, we also have 2n+2 divides dim ϕq−1 , so 2n+2 divides dim ϕ1 . We then see again using
Proposition 79.4 that a = 2n . In addition, dim ϕ ≡ 2a (mod 2n+2 ) so
(81.5) Sq2a
X (la+b+c−1 ) = 0
by Corollary 78.5 and Lemma 78.6.

Let Sym(ρ) = ρ + ρt for a cycle ρ on X̄ 2 be the symmetrization operation. The cycle
β is homogeneous, essential, symmetric, and does not contain any basis element having
hi with i < a + b as a factor. Consequently, we can write β = β0 + β1 with
³ ´
(81.6) β0 = Sym ha+b × la+b+c−1 ,
³P ´
(81.7) β1 = Sym hi+a+b × li+a+b+c−1
i∈I
for some set of positive integers I. Furthermore, since α does not contain any of hi × li
with i ∈ (0, a + b), we have
(81.8) µ = h0 × β + ha+b × γ + ν
for some essential cycle γ ∈ ChD+a+b+c−1 (X̄ 2 ) and some cycle ν ∈ Ch(X̄ 3 ) such that
the first factor of every basis element included in ν is of codimension > a + b. We can
decompose γ = γ0 + γ1 with
(81.9) γ0 = x · (h0 × la+b+c−1 ) + y · (la+b+c−1 × h0 ),
P P
(81.10) γ1 = hj × lj+a+b+c−1 + lj+a+b+c−1 × hj
j∈J j∈J 0
for some modulo 2 integers x, y ∈ Z/2Z and some sets of integers J, J 0 ⊂ (0, +∞).
We need the following:
Lemma 81.11. In the above, we have x = y = 1, I ⊂ [c, +∞), and J, J 0 ⊂ [a + b +
c, +∞).
Proof. To determine y, consider the cycle δ ∗ (µ) · (h0 × hc−1 ) ∈ Ch(X 2 ) where δ :
X 2 → X 3 is the morphism (x1 , x2 ) 7→ (x1 , x2 , x1 ). This rational cycle does not contain
h0 × l0 , while the coefficient of ha+b × la+b equals 1 + y. Consequently, y = 1 by the
minimality of α.
Similarly, using the morphism X 2 → X 3 , (x1 , x2 ) 7→ (x1 , x1 , x2 ) instead of δ, one
checks that x = 1 (although the value of x is not important for our future purposes).
To show that I ⊂ [c, +∞), assume to the contrary that 0 < i < c for some i ∈ I.
Then li+a+b ∈ Ch(XFq ) for this i and therefore the cycle
³ ´
0
li+a+b+c−1 = (pr 3 )∗ (l0 × li+a+b × h ) · µ
is Fq -rational, where pr 3 : X 3 → X is the projection onto the third factor. This contradicts
Corollary 72.6 because i + a + b + c − 1 ≥ a + b + c = jq (X) = i0 (XFq ).
81. ON THE 2-ADIC ORDER OF HIGHER WITT INDICES, I 353
To prove the statement for J, assume to the contrary that there exists a j ∈ J with
0 < j < a + b + c. Then lj ∈ Ch(XFq ) hence
³ ´
0
lj+a+b+c−1 = (pr 3 )∗ (la+b × lj × h ) · µ ∈ Ch(XFq ) ,
a contradiction. The statement for J 0 is proved similarly. ¤

Lemma 81.12. The cycle β is F1 -rational. The cycles γ and γ1 are Fq -rational.
Proof. Since F1 = F (X), the cycle β is F1 -rational by definition.
Let pr 23 : X 3 → X 2 , (x1 , x2 , x3 ) 7→ (x2 , x3 ) be the projection onto the product of
the second
¡ and the third ¢factors of X 3 . The cycle la+b is Fq -rational, therefore γ =
(pr 23 )∗ (la+b × h0 × h0 ) · µ is also Fq -rational. The cycle γ0 is Fq -rational as la+b+c−1 is
Fq -rational. It follows that γ1 is Fq -rational as well. ¤
Define
³ ¡ ¢´
ξ(χ) := ∗
δX 2 t∗12 (χ) ◦ Sq2a
X 3 (χ)
0 0
· (h × h × h c−a−1
) for any χ ∈ Ch(X̄ 3 ) .
We must prove that the cycle ξ(µ) · (h0 × hc−a−1 ) ∈ Ch(X 2 ) contains ha+b × la+b and does
not contain h0 × l0 , i.e., we have to show that ha+b × lb+c−1 ∈ ξ(µ) and h0 × lc−a−1 ∈
/ ξ(µ).
If h0 × lc−a−1 ∈ ξ(µ), then, passing from F to F1 = F (X), we have
¡ ¢
lc−a−1 = (pr 2 )∗ (l0 × h0 ) · ξ(µ) ∈ Ch(XF (X) ),
where pr 2 : X 2 → X is the projection onto the second factor of X 2 , contradicting Corollary
72.6 as c − a − 1 ≥ a = i1 (X) = i0 (XF (X) ).
It remains to show that ha+b × lb+c−1 ∈ ξ(µ). For any χ ∈ Ch(X̄ 2 ), write coeff(χ) ∈
Z/2Z for the coefficient of ha+b × lb+c−1 in χ. Since coeff(ν) = 0, it follows from (81.8)
that
¡ ¢ ¡ ¢
coeff ξ(µ) = coeff ξ(h0 × β + ha+b × γ) .
We claim that
¡ ¢ ¡ ¢
(81.13) coeff ξ(h0 × β) = 0 = coeff ξ(ha+b × γ) .
Indeed, since Sq2a 0 0 2a
X 3 (h × β) = h × SqX 2 (β) by Theorem 61.13, we have
³ ¡ 2a ¢´
0 0 ∗ 0 c−a−1
ξ(h × β) = h × δX β ◦ SqX 2 (β) · (h × h )
¡ ¢
where δX : X → X 2 is the diagonal morphism of X. Hence coeff ξ(h0 × β) = 0.
Since Sq2aX 3 (h
a+b
× γ) is ha+b × Sq2a j
X 2 (γ) plus terms having h with j > a + b as the
first factor by Remark 79.2, we have
¡ a+b ¢ ³ ³ ¡ 2a ¢´´
2a+2b ∗ 0 c−a−1
coeff ξ(h × γ) = coeff h × δX γ ◦ SqX 2 (γ) · (h × h ) =0.
This proves the claim.

It follows by claim (81.13) that
¡ ¢ ¡ ¢
(81.14) coeff ξ(µ) = coeff ξ(h0 × β + ha+b × γ) − ξ(h0 × β) − ξ(ha+b × γ) .
To compute the right hand side in (81.14), we need only the terms ha+b × Sq2a X 2 (γ) in the
formula for Sq2a
X 3 (h a+b
× γ) since the other terms do not effect coeff. Therefore, we see
that the right hand side coefficient in (81.14) is equal to
³ ³ ¡ ¢ ¡ 2a ¢´´
coeff ha+b × δX ∗
γ ◦ Sq2aX2 (β) · (h0
× h c−a−1
) + β ◦ Sq X2 (γ) · (h 0
× hc−a−1
) .
Consequently, to prove Proposition 81.4, it remains to prove
Lemma 81.15.
³ ¡ ¢ ¡ 2a ¢´
∗
δX γ ◦ Sq2a
X 2 (β) · (h0
× h c−a−1
) + β ◦ Sq X 2 (γ) · (h 0
× hc−a−1
) = lb+c−1 .
Proof. We start by showing that

³ ¡ ¢´
(81.16) ∗
δX β ◦ Sq2a
X 2 (γ) · (h 0
× hc−a−1
) =0.
Note that Sq2a vanishes on h0 × la+b+c−1 by relation (81.5). Therefore Sq2a (γ) = Sq2a (γ1 )
by (81.9). By Lemma 81.11 we may assume that dim X ≥ 4(a + b + c) − 2 (we shall need
this assumption in order to apply Proposition 81.1), otherwise γ1 = 0.
Looking at the exponent of the first factor of the basis elements contained in Sq2a (γ1 )
and using Lemma 81.11, we see that none ¡ of2athe basis 0elements h¢j ×lj+b+c−1 and lj+b+c−1 ×
j c−a−1
h with j < a + b + c are present in β ◦ Sq (γ1 ) · (h × h ) . As γ1 is Fq -rational by
Lemma 81.12, equation (81.16) holds by Proposition 81.1.
We compute Sq2a (β0 ) where β0 is as in (81.6). By Corollary 78.5 and Lemma 78.6,
we have Sq0 (ha+b ) = ha+b , Sqa (ha+b ) = h2a+b , and Sqj (ha+b ) = 0 for all others j ≤
2a. Moreover,
¡ 2a+b we have
¢ shown in (81.5) that Sq2a (la+b+c−1 ) = 0. Therefore, Sq2a (β0 ) =
Sym h × lb+c−1 by Theorem 61.13.
Using Lemma 81.11, we have
¡ ¢
γ0 ◦ Sq2a (β0 ) · (h0 × hc−a−1 ) = lb+c−1 × h0
and
³ ¡ ¢´
(81.17) ∗
δX γ0 ◦ Sq2a (β0 ) · (h0 × hc−a−1 ) = lb+c−1 .
¡ ¢
The composition γ0 ◦ Sq2a (β1 ) · (h0 × hc−a−1 ) is trivial. Indeed, by Lemma 81.11,
every basis element of the cycle Sq2a (β1 ) · (h0 × hc−a−1 ) has (as the second factor) either
lj with j ≥ 2a + b + c > 0 or hj with j ≥ b + 2c − 1 > a + b + c − 1, while the two basis
elements of γ0 have h0 and la+b+c−1 as the first factor. Consequently
³ ¡ 2a ¢´
∗ 0 c−a−1
(81.18) δX γ0 ◦ Sq (β1 ) · (h × h ) =0.
Looking at the exponent of the first factor of the basis elements contained in γ1 and
using Lemma 81.11, we see that none¡ of the basis elements h¢j × lj+b+c−1 and lj+b+c−1 × hj
with j < a + b + c is present in γ1 ◦ Sq2a (β) · (h0 × hc−a−1 ) . Therefore, the relation
³ ¡ ¢´
(81.19) ∗
δX γ1 ◦ Sq2a (β) · (h0 × hc−a−1 ) = 0
holds by Proposition 81.1 in view of Lemma 81.12.
Taking the sum of the relations in (81.16)–(81.19), we have established the proof of
Lemma 81.15. ¤
82. HOLES IN I n 355
This completes the proof of Proposition 81.4. ¤
Theorem 81.3 is proved. ¤

Corollary 81.20. Let ϕ be an anisotropic quadratic form over a field of characteristic
not two. If h = h(ϕ) > 1, then
¡ ¢
v2 (i1 ) ≥ min v2 (i2 ), . . . , v2 (ih ) − 1 .
Proof. For any odd-dimensional ϕ, the statement is trivial, as all iq are odd by
Corollary 79.6. Assume that the inequality fails for an even-dimensional anisotropic ϕ.
Note that in this case the difference
dim ϕ − i1 = i1 + 2(i2 + · · · + ih )
can not be a power of 2 because it is larger than 2n and not divisible by 2n+1 . Therefore,
by Corollary 80.10, the 1-primordial cycle on X 2 does produce an integer. Therefore, the
assumptions of Theorem 81.3 are satisfied, leading to a contradiction. ¤
Example 81.21. For an anisotropic quadratic form of dimension 6 and of trivial
discriminant, we have h = 2, i1 = 1, and i2 = 2. Therefore, the lower bound on v2 (i1 ) in
Corollary 81.20 is exact.
82. Holes in I n
Recall that F is a field of characteristic not two. For every integer n ≥ 1, we set
dim I n (F ) := {dim ϕ | ϕ ∈ I n (F ) and anisotropic} .
and
[
dim I n := dim I n (F ) ,
where the union is taken over all fields F (of characteristic 6= 2).
In this section, we determine the set dim I n . Theorem 82.8 states that dim I n is the
set of even non-negative integers without the following disjoint open intervals (which we
call holes in I n ):
Un−i = (2n+1 − 2i+1 , 2n+1 − 2i ) , i = n, n − 1, . . . , 1 .
The statement that U0 ∩ dim I n = ∅ is the Hauptsatz 23.8(1). This was originally proved
by Arason and Pfister [12, Hauptsatz]. The statement for U1 ∩ dim I n with n = 3 was
originally proved in 1966 by Pfister [114, Satz 14], with n = 4 in 1998 by Hoffmann [56,
Main Theorem], and for arbitrary n in 2000 by Vishik [140, Th. 6.4]. The statement
that Un−i ∩ dim I n = ∅ for any n and i was conjectured by Vishik [140, Conj. 6.5]. A
positive solution of the conjecture was announced by Vishik in 2002 but the proof is not
available; a proof was given in [74].
Proposition 82.1. Let ϕ be a nonzero anisotropic form of even dimension with
deg ϕ = n ≥ 1. If dim ϕ < 2n+1 then dim ϕ = 2n+1 − 2i+1 for some i ∈ [0, n − 1].
Proof. We use the notation of §81. We prove the statement by induction on h = h(ϕ).
The case of h = 1 is trivial.
So assume that h > 1. As dim ϕ1 < dim ϕ < 2n+1 and deg ϕ1 = deg ϕ, where ϕ1 is the
first (anisotropic) kernel form of ϕ, the induction hypothesis implies
dim ϕ1 = 2n+1 − 2i+1 for some i ∈ [1, n − 1].
Therefore, dim ϕ = 2n+1 − 2i+1 + 2i1 . Since dim ϕ < 2n+1 , we have i1 < 2i . In particular,
v2 (dim ϕ − i1 ) = v2 (i1 ). As i1 ≤ exp2 v2 (dim ϕ − i1 ) by Proposition 79.4, it follows that i1
is a 2-power, say i1 = 2j for some j ∈ [0, i − 1].
It follows by the induction hypothesis that each of the integers dim ϕ1 , . . . , dim ϕh is
divisible by 2i+1 . Therefore, v2 (iq ) ≥ i for all q ∈ [2, h] and hence by Corollary 81.20, we
have j ≥ i − 1. Consequently, j = i − 1, so dim ϕ = 2n+1 − 2i . ¤
Corollary 82.2. Let ϕ is an anisotropic quadratic form such that ϕ ∈ I n (F ) for
some n ≥ 1. If dim ϕ < 2n+1 then dim ϕ = 2n+1 − 2i+1 for some i ∈ [0, n].
Proof. We may assume that ϕ 6= 0. We have deg ϕ ≥ n by Corollary 25.12. Since
2deg ϕ ≤ dim ϕ < 2n+1 , we must have deg ϕ = n. The result follows from Proposition
82.1. ¤
Corollary 82.3. Let ϕ 6= 0 be an anisotropic quadratic form in I n (F ) with dim ϕ <
n+1
2 . Then the higher Witt indices of ϕ are the successive 2-powers:
i1 = 2i , i2 = 2i+1 , . . . , ih = 2n−1 ,
where i = log2 (2n+1 − dim ϕ) − 1 is an integer.
Proof. By Corollary 82.2, we have dim ϕ = 2n+1 − 2i+1 for i as in the statement of
Corollary 82.3, and dim ϕ1 = 2n+1 − 2j+1 for some j > i. It follows by Proposition 79.4
that i1 = 2i . We proceed by induction on dim ϕ. ¤
We now show that every even value of dim ϕ for ϕ ∈ I n (F ) not forbidden by Corollary
82.2 is possible over some F . We start with some preliminary work.
Lemma 82.4. Let ϕ be a nonzero anisotropic quadratic form in I n (F ) and dim ϕ <
2n+1 for some n ≥ 1. Then the 1-primordial cycle is the only primordial cycle in Ch(X 2 ).
Proof. We induct on h = h(ϕ). The case h = 1 is trivial, so we assume that h > 1.
Let pr 2∗ : Ch(X 2 ) → Ch(X12 ) be the homomorphism of Remark 72.5. Since the integer
dim ϕ − i1 lies inside the open interval (2n , 2n+1 ), it is not a 2-power. Hence by Corollary
80.10, we have pr 2∗ (π) 6= 0, where π ∈ Ch(X 2 ) is the 1-primordial cycle. Therefore, by
the induction hypothesis, the diagram of pr 2∗ (π) has points in every shell triangle. Thus,
the diagram of π itself has points in every shell triangle. By Theorem 73.28, this means
that π is the unique primordial cycle in Ch(X 2 ). ¤
Corollary 82.5. Let ϕ be a nonzero anisotropic quadratic form in I n (F ) and dim ϕ =
n+1
2 − 2 for some n ≥ 1. Then for any i > 0, the group ChD+i (X 2 ) contains no essential
element.
Proof. By Lemma 82.4, the 1-primordial cycle is the only primordial cycle in Ch(X 2 ).
Since i1 = 1 by Corollary 82.3, we have dim π = D. To finish we apply Theorem 73.28. ¤
82. HOLES IN I n 357
Lemma 82.6. Let F0 be a field (of char F0 6= 2),

F = F0 (t1j , t2j )1≤j≤n
the field of rational functions in 2n variables. Then the quadratic form
hht11 , . . . , t1n ii0 ⊥ − hht21 , . . . , t2n ii0
over F is anisotropic (where the prime denotes the pure subform of the Pfister form).
Proof. For any i = 0, 1, . . . , n, we set ϕi = hht11 , . . . , t1i ii and ψi = hht21 , . . . , t2i ii. We
prove that the form ϕ0i ⊥ − ψi0 is anisotropic by induction on i. For i = 0 the statement is
trivial. For i ≥ 1, we have:
0
ϕ0i ⊥ − ψi0 ' (ϕ0i−1 ⊥ − ψi−1 )⊥ t1i ϕi−1 ⊥ − t2i ψi−1 .
The summand ϕ0i−1 ⊥ − ψi−10
is anisotropic by the induction hypothesis, while the forms
ϕi−1 and ψi−1 are so by Corollary 19.6. Applying repeatedly Lemma 19.5 we conclude
that the whole form is anisotropic. ¤
Define the anisotropic pattern)anisotropic pattern of a quadratic form ϕ over F to be
the set of the integers dim(ϕK )an for all field extensions K/F . By Proposition 25.1, the
anisotropic pattern of a form ϕ coincides with the set
{dim ϕ − 2jq (ϕ) | q ∈ [0, h(ϕ)]}.
The following result is due to Vishik.
Proposition 82.7. Let F0 be a field (of char F0 6= 2) and integers n ≥ 1 and m ≥ 2.
Let
F = F0 (ti , tij )1≤i≤m, 1≤j≤n
the field of rational functions in variables ti and tij . Then the anisotropic pattern of the
quadratic form
ϕ = t1 · hht11 , . . . , t1n ii ⊥ . . . ⊥ tm · hhtm1 , . . . , tmn ii
over F is the set
¡ ¢
{2n+1 − 2i | i ∈ [1, n + 1] } ∪ 2Z ∩ [2n+1 , m · 2n ] .
Proof. We first show that all the integers 2n+1 − 2i are in the anisotropic pattern of
ϕ. Indeed, the anisotropic part of ϕ over the field E obtained from F by adjoining the
square roots of t31 , t41 , . . . , tm1 , of t1 and of −t2 , is isomorphic to the form
hht11 , . . . , t1n ii0 ⊥ − hht21 , . . . , t2n ii0
of dimension 2n+1 − 2. This form is anisotropic by Lemma 82.6. The anisotropic pattern
of this form is {2n+1 − 2i | i ∈ [1, n + 1] } by Corollary 82.3.
Now assume that there is an even integer in the interval [2n+1 , m · 2n ] not in the
anisotropic pattern of ϕ. Among all such integers take the smallest one and call it a. Let
b = a − 2 and c the smallest integer greater than a and lying in the anisotropic pattern
of ϕ. Let E be the field in the generic splitting tower of ϕ such that dim ψ = c where
ψ = (ϕE )an and Y the projective quadric given by the quadratic form ψ. Let π ∈ Ch(Y 2 )
be the 1-primordial cycle. We claim that
π = h0 × li1 −1 + li1 −1 × h0
where i1 = i1 (Y ). Indeed, since i1 = (c − b)/2 > 1 and iq (Y ) = 1 for all q such that
dim ψq ∈ [2n+1 − 2, b − 2], the diagram of the cycle π does not have any point in the
qth shell triangle for such q. For the integer q satisfying dim ψq = 2n+1 − 2, the cycle
pr 2∗ (π) ∈ Ch(Yq2 ) has dimension > dim Yq hence is 0 by Corollary 82.5. The relation
pr 2∗ (π) = 0 means that π has no point in any shell triangle with number > q.
It follows that π = h0 × li1 −1 + li1 −1 × h0 . By Corollary 80.8, the integer dim Y − i1 + 2
is a power of 2, say 2p . Since
dim Y − i1 + 2 = (c − 2) − (c − b)/2 + 2 = (b + c)/2 ,
the integer 2p lies inside the open interval (b, c). It follows that the integer 2p satisfies
2n+1 ≤ 2p < m · 2n and is not in the splitting pattern of the quadratic form ϕ. But
every integer ≤ m · 2n divisible by 2n is evidently in the anisotropic pattern of ϕ. This
contradiction establishes Proposition 82.7. ¤
Summarizing, we have
Theorem 82.8. For any integer n ≥ 1,
¡ ¢
dim I n = {2n+1 − 2i | i ∈ [1, n + 1]} ∪ 2Z ∩ [2n+1 , +∞) .
Proof. The inclusion ⊂ is given by Corollary 82.2, while the inclusion ⊃ follows by
Proposition 82.7. ¤
Remark 82.9. The case of dimension 2n+1 − 2i can be realized directly by difference
of two (i − 1)-linked n-fold Pfister forms (cf. Corollary 24.3).
83. On the 2-adic order of higher Witt indices, II

Throughout this section, X is an anisotropic quadric of dimension D over a field of
characteristic not two. We write i1 , . . . , ih and j1 , . . . , jh for the relative and absolute higher
Witt indices of X, respectively, where h is the height of X (cf. §81).
The main result of this section is Theorem 83.3. It is used to establish further relations
between higher Witt indices in Corollary 83.4.
We first establish further special properties of the 1-primordial cycle in addition to
those in Proposition 73.30 and Theorem 81.3.
Lemma 83.1. Let π ∈ Ch(X 2 ) be the 1-primordial cycle. Then SqjX 2 (π) = 0 for all
j ∈ (0, i1 ).
Proof. Let Sq = SqX 2 . Assume that Sqj (π) 6= 0 for some j ∈ (0, i1 ). By Remark
79.2, one sees that Sqj (π) has a non-trivial intersection with an appropriate jth order
derivative of π. As the derivative of π is minimal by Theorem 73.28, the cycle Sqj (π)
contains this derivative. It follows that Sqj (π) has a point in the first left shell triangle,
contradicting Lemma 79.3. ¤
Proposition 83.2. Let i be an integer such that hi ×l? is contained in the 1-primordial
cycle. Then i is divisible by 2n+1 for any n ≥ 0 satisfying i1 > 2n .
84. MINIMAL HEIGHT 359
Proof. Assume that the statement is false. Let i be the minimal integer not divisible
by 2n+1 and such that hi × l? is contained in the 1-primordial cycle π ∈ Ch(X 2 ).
Note that π contains only essential basis elements and is symmetric. As dim π =
D + i1 − 1, we have hi × li+i1 −1 ∈ π. ¡¢
For any non-negative integer k divisible by 2n+1 , the binomial coefficient kl with a
non-negative integer l is odd only if l is divisible by 2n+1 by Lemma 78.6. Therefore,
SqX (hk ) = hk (1 + h)k is a sum of powers of h with exponents divisible by 2n+1 . It follows
that the value SqjX (π) contains the element Sq0X (hi ) × SqjX (li+i1 −1 ) = hi × SqjX (li+i1 −1 ) for
any integer j. Since SqjX (π) = 0 for j ∈ (0, i1 ) by Lemma 83.1, we have
SqjX (li+i1 −1 ) = 0 for j ∈ (0, i1 ).
v2 (i)
Now look at the specific value Sq2X (li+i1 −1 ). Since i is not divisible by 2n+1 and i1 > 2n ,
the degree 2v2 (i) of the Steenrod operation lies in the interval (0, i1 ). By Corollary 78.5,
v2 (i)
the value Sq2X (li+i1 −1 ) is equal to li+i1 −1−2v2 (i) multiplied by the binomial coefficient
µ ¶
D − i − i1 + 2
.
2v2 (i)
The integer D − i1 + 2 = dim ϕ − i1 is divisible by 2n+1 by Proposition 79.4 as i1 > 2n .
Therefore the binomial coefficient is odd by Lemma 78.6. This is a contradiction so
establishes the result. ¤
Theorem 83.3. Let X be an anisotropic quadric over a field of characteristic not
two. Suppose that the 1-primordial cycle π ∈ Ch(X 2 ) produces the integer q. Then
v2 (iq ) ≥ v2 (i1 ).
Proof. Let n = v2 (i1 ). Then the integer 2n divides dim ϕ − i1 by Proposition 79.4.
Therefore, 2n divides dim ϕ as well.
We have hjq−1 × ljq−1 +i1 −1 ∈ π by definition of q. Consequently, by Proposition 83.2,
the integer jq−1 is divisible by 2n . It follows that 2n divides dim ϕq−1 = dim ϕ − 2jq−1 ,
where ϕq−1 is the (q − 1)th (anisotropic) kernel of ϕ. If m < n for m = v2 (iq ), then
applying Proposition 79.4, we have iq = i1 (ϕq−1 ) is equal to 2m and, in particular, smaller
than i1 . Therefore the 1-primordial cycle π has no points in the qth shell triangle. But
the point hjq−1 × ljq−1 +i1 −1 ∈ π lies in the qth shell triangle. This contradiction establishes
the theorem. ¤
¡ ¢
Corollary 83.4. We have v2 (i1 ) ≤ max v2 (i2 ), . . . , v2 (ih ) if the integer
dim ϕ − i1 = i1 + 2(i2 + · · · + ih )
is not a power of 2.
Proof. If the integer dim ϕ − i1 is not a 2-power then the 1-primordial cycle does
produce an integer by Corollary 80.10. The result follows by Theorem 83.3. ¤
84. Minimal height
Every non-negative integer n is uniquely representable in the form of an alternating
sum of 2-powers:
n = 2p0 − 2p1 + 2p2 − · · · + (−1)r−1 2pr−1 + (−1)r 2pr
for some integers p0 , p1 , . . . , pr satisfying p0 > p1 > · · · > pr−1 > pr + 1 > 0. We shall
write P (n) for the set {p0 , p1 , . . . , pr }. Note that pr coincides with the 2-adic order v2 (n)
of n. For n = 0 our representation is the empty sum so P (0) = ∅.
Define the height h(n) of the integer n as the number of positive elements in P (n).
So h(n) is the number |P (n)|, the cardinality of the set P (n) if n is even, while h(n) =
|P (n)| − 1 if n is odd.
In this section we prove the following theorem conjectured by Rehmann and originally
proved in [61]:
Theorem 84.1. Let ϕ be an anisotropic quadratic form over a field of characteristic
not two. Then
h(ϕ) ≥ h(dim ϕ) .
Remark 84.2. Let n ≥ 0 and ϕ anisotropic excellent quadratic form of dimension n.
It follows from Proposition 28.5 that h(ϕ) = h(n). Therefore, the bound in Theorem 84.1
is sharp.
We shall see (cf. Corollary 84.5) that Theorem 84.1 in odd dimensions is a consequence
of Proposition 79.4. In even dimensions we shall also need Theorem 81.3 and Theorem
83.3. We shall also need several combinatorial arguments to prove Theorem 84.1.
Suppose ϕ is anisotropic. Let ϕi be the ith (anisotropic) kernel form of ϕ, and ni =
dim ϕi , 0 ≤ i ≤ h(ϕ).
Lemma 84.3. For any i ∈ [1, h], the difference d(i) := h(ni−1 ) − h(ni ) satisfies the
following:
(I) If the dimension of ϕ is odd then |d(i)| = 1.
(II) If the dimension of ϕ is even then |d(i)| ≤ 2. Moreover,
(+2) If d(i) = 2 then P (ni ) ⊂ P (ni−1 ) and v2 (ni ) ≥ v2 (ni−1 ) + 2.
(+1) If d(i) = 1, the set difference P (ni ) \ P (ni−1 ) is either empty or consists of a
single element p, in which case both integers p − 1 and p + 1 lie in P (ni−1 ).
(0) If d(i) = 0, the set difference P (ni ) \ P (ni−1 ) consists of one element p and
either p − 1 or p + 1 lies in P (ni−1 ).
(-1) If d(i) = −1, the set difference P (ni )\P (ni−1 ) consists either of two elements
p − 1 and p + 1 for some p ∈ P (ni−1 ) or the set difference consists of one
element.
(-2) If d(i) = −2, the set difference P (ni ) \ P (ni−1 ) consists of two elements, i.e.,
P (ni ) ⊃ P (ni−1 ). Moreover, in this case one of these two elements is equal
to p + 1 for some p ∈ P (ni−1 ).
Proof. Write p0 , p1 , . . . , pr for the elements of P (ni−1 ) in descending order. We have
ni = ni−1 −2ii . We also know by Proposition 79.4, that there exists a non-negative integer
m such that 2m < ni−1 , ii ≡ ni−1 (mod 2m ), and 1 ≤ ii ≤ 2m . The condition 2m < ni−1
implies m < p0 . Let ps be the element with maximal even s satisfying m < ps .
If m = ps − 1 then ii = 2ps −1 − 2ps+1 + 2ps+2 − . . . and, therefore,
ni = 2p0 − 2p1 + · · · − 2ps−1 + 2ps+1 − 2ps+2 + · · · + (−1)r−1 2pr .
If s = r and pr−1 + 1 = pr−2 then P (ni ) equals P (ni−1 ) without pr−2 and pr . Otherwise,
P (ni ) equals P (ni−1 ) without ps .
So we may assume that m < ps − 1.

If s = r then ii = 2m and ni = ni−1 − 2m+1 . If m = pr − 2, we have P (ni ) is obtained
from P (ni−1 ) by replacing pr with pr − 1. If m < pr − 2, we have P (ni ) equals P (ni−1 )
with m + 1 added.
So we may assume in addition that s < r.
If ps − 1 > m > ps+1 then ii = 2m − 2ps+1 + 2ps+2 − . . . and, therefore,
ni = 2p0 − 2p1 + · · · − 2ps−1 + 2ps − 2m+1 + 2ps+1 − 2ps+2 + · · · + (−1)r+1 2pr .
This is the correct representation of ni and, therefore, P (ni ) equals P (ni−1 ) with m + 1
added.
It remains to consider the case with m ≤ ps+1 while s < r. In this case, first assume
that s = r − 1. Then ii = 2m and ni = ni−1 − 2m+1 .
If m < pr − 2 then P (ni ) equals P (ni−1 ) with pr + 1 and m + 1 added.
If m = pr − 2 then P (ni ) equals P (ni−1 ) with pr removed and pr + 1 and pr − 1 added.
If m = pr − 1, one has two possibilities. If pr−1 > pr + 2 then P (ni ) equals P (ni−1 )
with pr removed and pr + 1 added. If pr−1 = pr + 2 then P (ni ) equals P (ni−1 ) with pr
and pr−1 removed while pr + 1 added.
Finally, if m = pr then either pr−1 = pr + 2 and P (ni ) equals P (ni−1 ) without pr−1 ,
or P (ni ) equals P (ni−1 ) with pr + 2 added.
To finish the proof we may assume that m ≤ ps+1 and s < r − 1. We have: ii =
ps+2
2 − 2ps+3 + · · · + (−1)r 2pr and
ni = 2p0 − 2p1 + · · · + 2ps − 2ps+1 +1 + 2ps+1 − 2ps+2 + · · · + (−1)r+1 2pr .
So, if ps > ps+1 + 1 then P (ni ) equals P (ni−1 ) with ps+1 + 1 added; otherwise P (ni ) is
P (ni−1 ) with ps removed. ¤
Corollary 84.4. Let ϕ be an anisotropic odd-dimensional quadratic form and i ∈
[1, h(ϕ)]. Then
h(ni−1 ) − h(ni ) ≤ 1 .
Corollary 84.5. Let ϕ be an anisotropic quadratic form of odd dimension n. Then
h(ϕ) ≥ h(n).
Proof. Let h := h(ϕ). As dim ϕ is odd, nh = 1. Then h(nh ) = 0 and by Corollary
84.4, we have h(ni−1 ) − h(ni ) ≤ 1 for every i ∈ [1, h]. Therefore, h(n0 ) ≤ h. Since ϕ is
anisotropic, n = n0 , and the result follows. ¤
Remark 84.6. By Lemma 84.3, for any quadratic form ϕ of odd dimension n, we
have h(ni ) = h(ni−1 ) ± 1. Therefore h(ϕ) ≡ h(n) (mod 2).
Proposition 84.7. Let ϕ be an anisotropic quadratic form of even dimension n and
h := h(ϕ). Suppose that v2 (ni ) ≥ v2 (ni−1 ) + 2 for some i ∈ [1, h). Then the open interval
(i, h) contains an integer i0 satisfying |v2 (ni0 ) − v2 (ni−1 )| ≤ 1.
Proof. It suffices to consider the case i = 1. Note that h ≥ 2. Set p = v2 (n0 ).
By assumption, we have v2 (n1 ) ≥ p + 2. Therefore, v2 (i1 ) = p − 1. Clearly, the integer
n0 − i1 = i1 + n1 is not a power of 2. Therefore, by Corollary 80.10, the 1-primordial cycle
of Ch(X 2 ) produces an integer j ∈ [2, h]. We shall show that either v2 (nj−1 ) or v2 (nj )
lies in [p − 1, p + 1] for this j. We then take i0 = j − 1 in the first case and i0 = j in the
second case. Note that i0 6= 1 and i0 6= h as v2 (n1 ) ≥ p + 2, while v2 (nh ) = ∞.
By Theorem 83.3, we have v2 (ij ) ≥ p−1. Consequently, v2 (nj−1 ) ≥ p−1 by Proposition
79.4 as well. Since n1 = 2(i2 +· · ·+ij−1 )+nj−1 , it follows that v2 (i2 +· · ·+ij−1 )+1 ≥ p−1.
If v2 (i2 +· · ·+ij−1 ) < p+1 then v2 (nj−1 ) = v2 (i2 +· · ·+ij−1 )+1 ∈ [p−1, p+1]. So, we may
assume that v2 (i2 +· · ·+ij−1 ) ≥ p+1 and apply Theorem 81.3 to conclude that v2 (ij ) ≤ p.
We have v2 (ij ) ∈ {p − 1, p}. If v2 (nj−1 ) > v2 (ij ) + 1 then v2 (nj ) = v2 (ij ) + 1 ∈ {p, p + 1}.
If v2 (nj−1 ) = v2 (ij ) + 1 then v2 (nj−1 ) ∈ {p, p + 1}. Finally, If v2 (nj−1 ) < v2 (ij ) + 1 then
v2 (nj−1 ) = v2 (ij ), hence v2 (nj−1 ) ∈ {p − 1, p}. ¤
Corollary 84.8. Let ϕ be an anisotropic quadratic form of even dimension n. Sup-
pose that v2 (ni ) ≥ v2 (ni−1 ) + 2 for some i ∈ [1, h). Set p = v2 (ni−1 ). Then there exists
i0 ∈ (i, h) such that the set P (ni0 ) contains an element p0 with |p0 − p| ≤ 1.
Proof. Let i0 be the integer in the conclusion of Proposition 84.7. Then p0 = v2 (ni0 )
works. ¤
We now prove Theorem 84.1.
Proof of Theorem 84.1. By Corollary 84.5, we need only to prove Theorem 84.1
for even-dimensional forms. Let h := h(ϕ) ≥ 1 and {n0 > n1 > · · · > nh } with ni = dim ϕi
be the anisotropic pattern of ϕ with n = n0 even.
Let H be the set {1, 2, . . . , h}. For any i ∈ H, let d(i) := h(ni−1 ) − h(ni ). Recall
that d(i) ≤ 2 for any i ∈ H by Lemma 84.3. Let C be the subset of H consisting
of all those i ∈ H such that d(i) = 2. We shall construct a map f : C → H satisfying
d(j) ≤ 1−|f −1 (j)| for any j ∈ f (C). In particular, we shall have f (C) ⊂ H \C. Once such
a map is constructed, we establish Theorem 84.1 as follows: The subsets f −1 (j)∪{j} ⊂ H,
where j runs over H \ C, are disjoint and cover¡ P H. In addition,
¢ the average value of d on
each such subset is ≤ 1, so the average value i∈H d(i) /h = h(n)/h of d on H is ≤ 1,
i.e., h(n) ≤ h.
So it remains to define the map f with the desired properties. Let i ∈ C. By Lemma
84.3, we have v2 (ni ) ≥ v2 (ni−1 ) + 2. Therefore, by Corollary 84.8, there exists i0 ∈ (i, h)
such that the set P (ni0 ) contains an element p0 satisfying |p0 − p| ≤ 1 for p = v2 (ni−1 ).
Taking the minimal i0 with this property, set f (i) = i0 . We also define g(i) to be the
minimal element of P (nf (i) ) satisfying |g(i) − p| ≤ 1.
This defines the map f . To finish, we must show that f has the desired property.
First observe that by the definition of f , for any i ∈ C and any j ∈ [i, f (i) − 1]
the set P (nj ) does not contain any element p with |p − v2 (ni−1 )| ≤ 1. It follows that
if f (i1 ) = f (i2 ) for some i1 6= i2 then for p1 = v2 (ni1 −1 ) with p2 = v2 (ni2 −1 ) one has
|p2 − p1 | ≥ 2. Moreover, if g(i1 ) = g(i2 ) then |p1 − p| ≤ 1 by definition of g and
|p2 − p| ≤ 1 for p = g(i1 ) = g(i2 ). Therefore, we have
(84.9) if f (i1 ) = f (i2 ) and g(i1 ) = g(i2 ) for some i1 6= i2
then |p2 − p1 | = 2 for p1 = v2 (ni1 −1 ) and p2 = v2 (ni2 −1 ).
Let j ∈ f (C). By the definition of f , the set difference P (nj ) \ P (nj−1 ) is non-empty.
Then d(j) 6= 2 by Lemma 84.3(II+2). Moreover, the above set difference contains an
element p satisfying {p − 1, p + 1} 6⊂ P (nj−1 ). Consequently, d(j) 6= 1 by Lemma

84.3(II+1). Therefore, d(j) ≤ 0 by Lemma 84.3.
Now let j be an element of f (C) with |f −1 (j)| ≥ 2. Let i1 < i2 be two different
elements of f −1 (j). Note that i1 < i2 < j. Moreover, if p1 = v2 (ni1 −1 ) and p2 = v2 (ni2 −1 )
then, by the definition of f (i1 ), we have |p2 − p1 | > 1. We shall show that d(j) ≤ −1.
We already know that d(j) ≤ 0. If d(j) = 0, then by Lemma 84.3(II-0), the set difference
P (nj ) \ P (nj−1 ) consists of one element p0 and either p0 − 1 or p0 + 1 lies in P (nj−1 ). Since
the difference P (nj ) \ P (nj−1 ) consists of one element p0 , we have p0 = g(i1 ) = g(i2 ). It
follows that {p1 , p2 } = {p0 − 1, p0 + 1}. Consequently, the set P (nj−1 ) contains neither
p0 − 1 nor p0 + 1, a contradiction. Thus we have proved that d(j) ≤ −1 if |f −1 (j)| ≥ 2.
Now let j be an element of f (C) with |f −1 (j)| ≥ 3. Let i1 , i2 , i3 be three different
elements of f −1 (j). The equalities g(i1 ) = g(i2 ) = g(i3 ) cannot take place simultaneously,
as otherwise, by (84.9), we would have |p2 − p1 | = 2, |p3 − p2 | = 2, and |p1 − p3 | = 2, a
contradiction. However, the set difference P (nj ) \ P (nj−1 ) can have at most two elements.
Therefore, we may assume that g(i1 ) = g(i2 ) and that g(i3 ) is different from g(i1 ) = g(i2 ).
Set p0 = g(i1 ) = g(i2 ). We shall show that d(j) = −2. We already know that d(j) ≤ −1.
If d(j) = −1 then by Lemma 84.3(II-1), the set difference P (nj ) \ P (nj−1 ) consists of p̃ − 1
and p̃ + 1 for some p̃ ∈ P (nj−1 ). However, p0 is neither p̃ − 1 nor p̃ + 1, a contradiction.
We finish the proof by showing that |f −1 (j)| is never ≥ 4. Indeed, if |f −1 (j)| ≥ 4,
then the set difference P (nj ) \ P (nj−1 ) contains two elements p0 and p00 with none of p0 ± 1
or p00 ± 1 lying in P (nj−1 ), contradicting Lemma 84.3. ¤
CHAPTER XVI
The variety of maximal totally isotropic subspaces
So far in this book, the projective quadric has been the only variety associated with
a quadratic form. In this chapter, we introduce another variety, the variety of maximal
isotropic subspaces.
85. The variety Gr(ϕ)

Let ϕ be a non-degenerate quadratic form on V over F . In this chapter, we study the
scheme Gr(ϕ) of maximal totally isotropic subspaces of V . We view Gr(ϕ) as a closed
subscheme of the Grassmannian variety of V . Let n be the integer part of (dim ϕ − 1)/2,
so dim ϕ = 2n + 1 or 2n + 2.
Example 85.1. If dim ϕ = 1, we have Gr(ϕ) = Spec F . If dim ϕ = 2 or 3 then Gr(ϕ)
coincides with the quadric of ϕ, that is Gr(ϕ) = Spec C0 (ϕ) if dim ϕ = 2 and Gr(ϕ) is the
conic curve associated to the quaternion algebra C0 (ϕ) if dim ϕ = 3.
The orthogonal algebraic group O(V, ϕ) acts transitively on Gr(ϕ). Let O+ (V, ϕ) be
the (connected) special orthogonal group (cf. [86, §23]). If dim ϕ is odd, O+ (V, ϕ) acts
transitively on Gr(ϕ) and therefore, Gr(ϕ) is a smooth projective variety over F .
Suppose that dim ϕ = 2n + 2 is even. Then the group O(V, ϕ) has two connected
components, one is O+ (V, ϕ). The factor group O(V, ϕ)/ O+ (V, ϕ) is identified with the
Galois group over F of the center Z of the even Clifford algebra C0 (V, ϕ). Recall that Z
is an étale quadratic F -algebra (cf. Proposition 11.6). The class of Z(ϕ) in Ét2 (F ) is the
discriminant of ϕ (cf. §13).
A point of Gr(ϕ) over a commutative ring R is a totally isotropic direct summand
P of rank n + 1 of the R-module VR = V ⊗F R. Since p2 = 0 in the Clifford algebra
C(V, ϕ)R for every pV∈ P , the inclusion of P into VR gives rise to an injective R-module
homomorphism h : n+1 P → C(V, ϕ)R . Let W be the image of h. Since ZW = W , left
multiplication by elements of the center Z of C0 (V, ϕ) defines an F -algebra homomorphism
Z → EndR (W ) = R. Therefore, we have a morphism Gr(ϕ) → Spec Z, so Gr(ϕ) is a
scheme over Z.
If the discriminant of ϕ is trivial, i.e., Z = F × F , the scheme Gr(ϕ) has two smooth
(irreducible) connected components Gr1 and Gr2 permuted by O(V, ϕ)/ O+ (V, ϕ). More
precisely, they are isomorphic under any reflection of V . If Z is a field, the discriminant
of ϕZ is trivial and therefore Gr(ϕ) is isomorphic to a connected component of Gr(ϕZ ).
The varieties of even and odd dimensional forms are related by the following statement.
Proposition 85.2. Let ϕ be a non-degenerate quadratic form on V over F of di-
mension 2n + 2 and trivial discriminant and ϕ0 a non-degenerate subform of ϕ on a
365
366 XVI. THE VARIETY OF MAXIMAL TOTALLY ISOTROPIC SUBSPACES
subspace V 0 ⊂ V of codimension 1. Let Gr1 be a connected component of Gr(ϕ). Then

∼
the assignment U 7→ U ∩ V 0 gives rise to an isomorphism Gr1 → Gr(ϕ0 ).
Proof. Since both varieties Gr1 and Gr(ϕ0 ) are smooth, it suffices to show that the
assignment induces a bijection on points over any field extension L/F . Moreover, we may
assume that L = F . Let U 0 ⊂ V 0 be a totally isotropic subspace of dimension n. Then the
orthogonal complement U 0 ⊥ of U 0 in V is (n + 2)-dimensional and the induced quadratic
form on H = U 0 ⊥ /U1 has trivial discriminant (i.e., H is a hyperbolic plane). The space
H has exactly two isotropic lines permuted by a reflection. Therefore, the pre-images of
these lines in V are two totally isotropic subspaces of dimension n + 1 living in different
components of Gr(ϕ). Thus exactly one of them represents a point of Gr1 over F . ¤
Let ϕ0 be a non-degenerate subform of codimension 1 of a non-degenerate quadratic
form ϕ of even dimension. Let Z be the discriminant of ϕ. By Proposition 85.2, we have
Gr(ϕ0 )Z is isomorphic to a connected component Gr1 of Gr(ϕZ ) and therefore, Gr(ϕ) '
Gr1 ' Gr(ϕ0 )Z .
Example 85.3. If dim ϕ = 4 then Gr(ϕ) is the conic curve (over Z) associated to the
quaternion algebra C0 (ϕ).
Exercise 85.4. Show that if 3 ≤ dim ϕ ≤ 6 then Gr(ϕ) is isomorphic to the Severi-
Brauer variety associated to the even Clifford algebra C0 (ϕ).
86. The Chow ring of Gr(ϕ) in the split case

In this section, we present the calculation of the Chow groups of Gr(ϕ) given by
Vishik in [139]. Set Gr := Gr(ϕ) and r := dim ϕ − n − 1, where n is the integer part
of (dim ϕ − 1)/2. We define the tautological vector bundle E over Gr of rank r to be
the restriction of the tautological vector bundle over the Grassmannian variety of V , i.e.,
variety E is the closed subvariety of the trivial bundle V 1 := V × Gr consisting of pairs
(u, U ) such that u ∈ U . The projective bundle P(E) is a closed subvariety of X × Gr,
where X is the (smooth) projective quadric of ϕ.
Let E ⊥ be the kernel of the natural morphism V 1 → E ∨ given by the polar bilinear
form bϕ . If dim ϕ = 2n + 2, we have U ⊥ = U for any totally isotropic subspace U ⊂ V of
dimension n + 1, hence E ⊥ = E.
Suppose that dim ϕ = 2n + 1. For any totally isotropic subspace U ⊂ V of dimension
n, the orthogonal complement U ⊥ contains U as a subspace of codimension 1. Therefore,
E ⊥ is a vector bundle over Gr of rank n + 1 containing E. The fiber of E ⊥ over U is the
orthogonal complement U ⊥ .
Suppose in addition that ϕ is isotropic. Choose an isotropic line L ⊂ V . Set Ve = L⊥ /L
and let ϕ̃ be the quadratic form on Ve induced by ϕ, Denote the projective quadric of ϕ̃
e
by X.
A totally isotropic subspace of Ve of dimension r − 1 is of the form U/L, where U is a
totally isotropic subspace of V of dimension n containing L. Therefore, we can view the
f := Gr(ϕ̃) of maximal totally isotropic subspaces of Ve as a closed subvariety of
variety Gr
Gr. Denote by i : Grf → Gr the closed embedding.
86. THE CHOW RING OF Gr(ϕ) IN THE SPLIT CASE 367
Let U be a totally isotropic ¡subspace of V ¢ of dimension r that does not contain L.

Then dim(U ∩ L⊥ ) = r − 1 and (U ∩ L⊥ ) + L /L is a totally isotropic subspace of Ve of
dimension r − 1.
¡ ¢
Lemma 86.1. The morphism f : Gr \Gr f → Gr f that takes U to (U ∩ L⊥ ) + L /L is
an affine bundle.
Proof. We use the criterion of Lemma 52.12. Let R be a local commutative F -
algebra. An F -morphism Spec R → Gr, f or equivalently an R-point of Gr, f is given by a
e
direct summand U of the R-module VR = V ⊗F R of rank r with LR ⊂ U e.
An R-point of Gr \Gr f is a totally isotropic direct summand U of VR of rank r with
U + LR a direct summand U of VR of rank r + 1. If f (U ) = U e /LR then U
e ⊂ U + LR . The
assignment U 7→ (U + LR )/U e gives rise to an isomorphism between the fiber Spec R × f
Gr
f of f over U
(Gr \Gr) e /LR and PR (VR /Ue ) \ PR (L⊥ /U
e ) ' An . By Lemma 52.12, we have f
R R
is an affine bundle. ¤
Note that dim Gr = dim Grf + n, so dim Gr = n(n + 1)/2.
By Lemma 86.1, Gr is a cellular variety with the short filtration Gr f ⊂ Gr, hence by
Corollary 66.4, we have a decomposition of Chow motives:
(86.2) M (Gr) = M (Gr) f ⊕ M (Gr)(n).
f
The morphism M (Gr) f → M (Gr) is induced by the embedding i : Gr f → Gr and the

f
morphism M (Gr)(n) → M (Gr) is given by the transpose of the closure of the graph of f ,
the class of which we shall denote by β ∈ CH(Gr f × Gr).
For the rest of this section, we shall assume that ϕ is split. It follows by induction,
using (86.2) and Example 85.1, that CH(Gr) is a free abelian group of rank 2r+1 . We
shall determine multiplicative structure of CH(Gr).
Since the motive of X (and also Gr) is split, we have CH(X × Gr) = CH(X) ⊗ CH(Gr)
by Proposition 64.3. In other words, CH(X × Gr) is a free module over CH(Gr) with basis
{hk × [Gr], lk × [Gr] | k ∈ [0, n − 1]} if dim ϕ = 2n + 1,
k
{h × [Gr], lk × [Gr], ln , ln0 | k ∈ [0, n − 1]} if dim ϕ = 2n + 2.
Note that in the even dimensional case, we have assumed that X is oriented.
In both cases, P(E) is a closed subvariety of X × Gr of codimension n. Therefore, in
the odd dimensional case there are unique elements ek ∈ CHk (Gr), k ∈ [0, n], satisfying
Xn
(86.3) [P(E)] = ln−1 × e0 + hn−k × ek
k=1
in CH(X × Gr). Pulling this back with respect to the canonical morphism XF (Gr) →
X × Gr, we see that e0 = 1.
In the even dimensional case, there are unique elements ek ∈ CHk (Gr), k ∈ [0, n]
and e00 ∈ CH0 (Gr), satisfying
n
X
(86.4) [P(E)] = ln × e0 + ln × e0 +
0 0
hn−k × ek
k=1
in CH(X × Gr). Choose a totally isotropic subspace U ⊂ V of dimension n + 1 so that

[P(U )] = ln in CH(X) and let U 0 be a reflection of U . It follows from Exercise 68.4 that
[P(U 0 )] = ln0 . Let g denote the generic point of Gr whose closure contains [U ]. Let g 0 be
another generic point of Gr whose closure contains [U 0 ]. Note that CH0 (Gr) = Z[g]⊕ Z[g 0 ].
Pulling back equation (86.4) with respect to the two morphisms X → X × Gr given by
the points [U ] and [U 0 ] respectively, we see that e0 = [g] and e00 = [g 0 ]. In particular, e0
and e00 are orthogonal idempotents of CH0 (Gr) hence e0 + e00 = 1.
It follows that for every totally isotropic subspace W ⊂ V of dimension n + 1 with
[W ] in the closure of g (resp. g 0 ), we have [W ] = ln (respectively [W ] = ln0 ). In particular,
to give an orientation of X is to choose one of the two connected components of Gr.
The multiplication rule in Proposition 68.1 for CH(X) implies that in both cases
¡ ¢
ek = p∗ (ln−k × 1) · [P(E)]
for k ∈ [1, n], where p : X × Gr → Gr is the projection.
We view the cycle γ = [P(E)] in CH(X ×Gr) as the incidence correspondence X Ã Gr.
It follows from Proposition 63.2 that the induced homomorphism γ∗ : CH(X) → CH(Gr)
takes ln−k to ek .
Let s : P(E) → Gr and t : P(E) → X be the two projections. Proposition 62.6
provides the following simple formula for ek :
(86.5) ek = s∗ ◦ t∗ (ln−k ).
f in CHn (Gr).
Lemma 86.6. We have en = [Gr]
f ∩
Proof. The element t∗ (l0 ) coincides with the cycle of the intersection ([L] × Gr)
P(E). It follows from (86.5) that [Gr]
f = s∗ ◦ t (l0 ) = en .
∗
¤
We write h̃ and ˜li for the standard generators of CH(X). e Recall that the incidence
correspondence α : Xe Ã X is given by the schemes of pairs (A/L, B) of one-dimensional
isotropic subspaces of Ve and V respectively with B ⊂ A. By Lemma 72.3, we can orient
Xe (in the case dim ϕ is even) so that α∗ (˜lk−1 ) = lk and α∗ (lk ) = ˜lk−1 for all k.
f the elements given by (86.3) or (86.4) for Gr.
Denote by ẽk ∈ CHk (Gr) f Similarly, we
have the incidence correspondence γ̃ : Xe Ã Gr f with γ̃∗ (˜ln−1−k ) = ẽk .
Lemma 86.7. The diagram of correspondences

α α t
e /o o/ o/ / X o/ /o o/ / X
X e
O O O
γ̃ O γ O γ̃ O
O O O
² β ² ²
it/o
f /o /o o/ / Gr /o
Gr /o / f
Gr
is commutative.
Proof. By Corollary 57.21, all calculations can be done on the level of cycles rep-
resenting the correspondences. By definition of the composition of correspondences, the
e × Gr consisting
compositions γ ◦ α and β ◦ γ̃ coincide with the cycle of the subscheme of X
of all pairs (A/L, U ) with dim(A + U ) ≤ n + 1. Similarly, the compositions γ̃ ◦ αt and
86. THE CHOW RING OF Gr(ϕ) IN THE SPLIT CASE 369
f consisting of all pairs (B, U

it ◦ γ coincide with the cycle of the subscheme of X × Gr e /L)
with B ⊂ U e. ¤
Corollary 86.8. We have β∗ (ẽk ) = ek and i∗ (ek ) = ẽk for all k ∈ [0, n − 1].
Proof. The equalities β∗ (ẽ0 ) = e0 and i∗ (e0 ) = ẽ0 follow from the fact that X and
e have compatible orientations. If k ≥ 1, we have by Lemma 86.7,
X
β∗ (ẽk ) = β∗ ◦ γ̃∗ (l0 ) = γ∗ ◦ α∗ (˜ln−1−k ) = γ∗ (ln−k ) = ek ,
n−1−k
∗
i (ek ) = it∗ (ek ) = it∗ ◦ γ∗ (ln−k ) = γ̃∗ ◦ α∗t (ln−k ) = γ̃∗ ◦ α∗ (ln−k ) = γ̃(ln−1−k ) = ẽk . ¤
For a subset I of [0, n] let eI be the product of ek for all k ∈ I. Similarly, we define
ẽJ for any subset J ⊂ [0, n − 1].
Corollary 86.9. We have i∗ (ẽJ ) = eJ · en = eJ∪{n} for every J ⊂ [0, n − 1].
Proof. By Corollary 86.8, we have i∗ (eJ ) = ẽJ . It follows from Lemma 86.6 and the
Projection Formula 56.9 that
i∗ (ẽJ ) = i∗ (i∗ (eJ ) · 1) = eJ · i∗ (1) = eJ · en = eJ∪{n} . ¤
Corollary 86.10. The monomial e[0, n] = e0 e1 · · · en is the class of a rational point
in CH0 (Gr).
Proof. The statement follows from the formula e[0, n] = i∗ (ẽ[0, n−1] ) and by induction
on n. ¤
f → Gr be the open embedding. Let f : Gr \Gr
Let j : Gr \Gr f → Gr f be the morphism
in Lemma 86.1.
Lemma 86.11. We have f ∗ (ẽJ ) = j ∗ (eJ ) for any J ⊂ [0, n − 1].
Proof. It suffices to prove that f ∗ (ẽk ) = j ∗ (ek ) for all k ∈ [0, n − 1]. By the
construction of β (cf. §66), we have β t ◦ j = f . It follows from Corollary 86.8 that
∗
f ∗ (ẽk ) = j ∗ ◦ (β t ) (ẽk ) = j ∗ ◦ β∗ (ẽk ) = j ∗ (ek ). ¤
Theorem 86.12. Let ϕ be a non-degenerate quadratic form on V over F of dimension
2n + 1 or 2n + 2 and Gr the scheme of maximal totally isotropic subspaces of V . Then
the set of monomials eI for all 2n subsets I ⊂ [1, n] is a basis of CH(Gr) over CH0 (Gr).
Proof. We induct on n. The localization property gives the exact sequence (cf. the
proof of Theorem 66.2)
j ∗
f − i∗ f → 0.
0 → CH(Gr) → CH(Gr) −
→ CH(Gr \Gr)
By the induction hypothesis and Corollary 86.9, the set of monomials eI for all I containing
f → CH(Gr \Gr)
n is a basis of the image of i∗ . Since f ∗ : CH(Gr) f is an isomorphism by
Theorem 52.13, again by the induction hypothesis and Lemma 86.11, the set of all the
elements j ∗ (eI ) with n ∈ f The statement follows.
/ I is a basis of CH(Gr \Gr). ¤
We now can compute the Chern classes of the tautological vector bundle E over Gr.
Proposition 86.13. We have ck (V 1/E) = ck (E ∨ ) = 2ek and ck (E) = (−1)k 2ek for
all k ∈ [1, n].
Proof. Let s : X → P(V ) be the closed embedding. Let H denote the class of a
hyperplane in P(V ). We have s∗ (hk ) = 2H k+1 for all k ≥ 0.
First suppose that dim ϕ = 2n + 1. It follows from (86.3) that
n
X
[P(E)] = s∗ (ln−1 ) × 1 + 2H n+1−k × ek
k=1
¡ ¢
in CH P(V ) × Gr .
On the other hand, by Proposition 58.10, applied to the subbundle E of V , we have
n+1
X
[P(E)] = H n+1−k × ck (V 1/E).
k=0
It follows from the Projective Bundle Theorem 53.10 that ck (V 1/E) = 2ek for k ∈ [1, n].
By duality, V 1/E ' (E ⊥ )∨ . Note that the line bundle E ⊥ /E carries a non-degenerate
quadratic form, hence is isomorphic to its dual. Since
¡ ⊥Pic(Gr)
¢ = CH1 (Gr) is torsion free,
we conclude that E /E ' 1. Therefore, c(E ) = c (E ) = c(V 1/E). The last equality
⊥ ∨ ∨
follows from Example 58.7.

The proof in the case dim ϕ = 2n+2 proceeds along similar lines: one uses the equality
(86.4) and the duality isomorphism V 1/E ' E ∨ . ¤
Remark 86.14. Let ϕ be a non-degenerate quadratic form that is not necessarily split.
By Proposition 86.13, the classes 2ek , k ≥ 1, that are a priori defined over a splitting field
of ϕ, are in fact defined over F .
In order to determine the multiplicative structure of CH(Gr), we present the set of
defining relations between the ek . For convenience, we set ek = 0 if k > n.
Since c(V 1/E) · c(E) = c(V 1) = 1 and CH(Gr) is torsion free, it follows from Propo-
sition 86.13 that
(86.15) e2k − 2ek−1 ek+1 + 2ek−2 ek+2 − · · · + (−1)k−1 2e1 e2k−1 + (−1)k e2k = 0
for all k ≥ 1.
Proposition 86.16. The equalities (86.15) form a set of defining relations between
the generators ek of the ring CH(Gr) over CH0 (Gr).
Proof. Let A be the factor ring of the polynomial ring Z[t1 , t2 , . . . , tn ] modulo the
ideal generated by polynomials giving the relations (86.15). We claim that the ring
homomorphism A → CH(Gr) taking tk to ek is an isomorphism.
Call a monomial tr11 tr22 · · · trnn with ri ≥ 0 basic if rk = 0 or 1 for every k. By Theorem
86.12, it is sufficient to prove that the ring A is generated by classes of basic monomials.
We define the weight w(m) of a monomial m = tr11 tr22 · · · trnn by the formula
n
X
w(m) = k 2 · rk
k=1
and the weight of a polynomial f (t1 , . . . , tn ) over Z as the minimum of weights of its
non-zero monomials. Clearly, w(m · m0 ) = w(m) + w(m0 ). For example, in the formula
87. THE CHOW RING OF Gr(ϕ) IN THE GENERAL CASE 371
(86.15), we have w(t2k ) = 2k 2 , w(tk−i tk+i ) = 2k 2 + 2i2 , and w(t2k ) = 4k 2 . Thus, t2k is the
monomial of the lowest weight in the formula (86.15).
Let f be a polynomial representing an element of the ring A. Applying formula (86.15)
to the square of a variable tk in a non-basic monomials of f of the lowest weight increases
the weight but not the degree of f . Since the weight of a polynomial of degree d is at
most n2 d, we eventually reduce to a polynomial having only basic monomials. ¤
The relations (86.15) look particularly simple modulo 2:
e2k ≡ e2k mod 2 CH(Gr) for all k ≥ 1.
Proposition 86.17. Let ϕ be a split non-degenerate quadratic form on V over F
of dimension 2n + 2 and ϕ0 a non-degenerate subform of ϕ on a subspace V 0 ⊂ V of
0
codimension 1. Let f denote the morphism ¡Gr(ϕ) → ¢ Gr(ϕ ) taking U to U ∩ V 0 , and e0k ,
0 ∗ 0
k ≥ 1, denote the standard generators of CH Gr(ϕ ) . Then f (ek ) = ek for all k ∈ [1, n].
Proof. Denote by E → Gr(ϕ) and E 0 → Gr(ϕ0 ) the tautological vector bundles of
ranks n + 1 and n respectively. The line bundle
E/f ∗ (E 0 ) = E/(V 0 1 ∩ E) ' (E + V 0 1)/V 0 1 = V 1/V 0 1
is trivial. In particular, c(E) = c(f ∗ E 0 ) = f ∗ c(E 0 ). It follows from Proposition 86.13 that
¡ ¢ ¡ ¢
2f ∗ (e0k ) = (−1)k f ∗ ck (E 0 ) = (−1)k f ∗ ck (f ∗ E 0 ) = (−1)k ck (E) = 2ek .
¡ ¢
The result follows since CH Gr(ϕ) is torsion free. ¤
87. The Chow ring of Gr(ϕ) in the general case

Let ϕ be a non-degenerate quadratic form of dimension 2n + 1 on V over an arbitrary
field F and X = Xϕ . Let Y be a smooth proper scheme over F and h : Y → Gr = Gr(ϕ)
a morphism. We set E 0 = h∗ (E), where E is the tautological vector bundle over Gr, and
view P(E 0 ) as a closed subscheme of X × Y .
Proposition 87.1. The CH(Y )-module CH(X × Y ) is free with basis hk , hk · [P(E 0 ]
with k ∈ [1, n − 1].
Proof. Let V 1 denote the trivial vector bundle V × Y over Y . We claim that
the restriction f : T = (X × Y ) \ P(E 0 ) → P(V 1/E 0 ⊥ ) of the natural morphism f :
P(V 1) \ P(E 0 ⊥ ) → P(V 1/E 0 ⊥ ) is an affine bundle. To do so we use the criterion of
Lemma 52.12.
Let R be a local commutative F -algebra. An F -morphism Spec R → P(V 1/E 0 ⊥ ),
equivalently, an R-point of P(V 1/E 0 ⊥ ) determines a pair (U, W ) where U is a totally
isotropic direct summand of VR of rank n and W is a direct summand of VR of rank n + 2
containing U ⊥ . Since rank W ⊥ = n − 1, one can choose an R-basis of W so that the
restriction of the quadratic form ϕ on W is equal to xy + az 2 for some a ∈ R× and U
is given by x = z = 0 in W . Therefore, the fiber Spec R ×P(V 1/E 0 ⊥ ) T is given by the
equation y/x = a(z/x)2 over R and hence is isomorphic to an affine space over R. By
Lemma 52.12, f is an affine bundle.
Thus X × Y is equipped with the structure of a cellular scheme. In particular, we

have a (split) exact sequence
¡ ¢ i∗
0 → CH P(E 0 ) −→ CH(X × Y ) → CH(T ) → 0
and the isomorphism
¡ ⊥ ¢ ∼
f ∗ : CH P(V 1/E 0 ) → CH(T ).
¡ ¢
The restriction of the canonical line bundle over P(V ) to X × Y and CH P(E 0 ) are
also canonical bundles. It follows from the Projective Bundle Theorem 53.10 and the
Projection Formula 56.9 that the image of i∗ is a free CH(Y )-module with basis hk ·[P(E 0 )],
k ∈ [0, n − 1].
The geometric description of the canonical line bundle given in §104.C shows that
the pull-back with respect to f of the canonical line bundle is the restriction to T of the
canonical bundle on X × Y . Again, it follows from the Projective Bundle Theorem 53.10
that CH(T ) is a free CH(Y )-module with basis the restrictions of hk , k ∈ [0, n − 1], on
T . The statement readily follows. ¤
Remark 87.2. The proof of Proposition 87.1 gives the motivic decomposition
¡ ¢ ¡ ⊥ ¢
M (X × Y ) = M P(E 0 ) ⊕ M P(V 1/E 0 ) (n).
As in the case of quadrics, we write CH(Gr) for the colimit of CH(GrL ) over all field
extensions L/F and CH(Gr) for the image of CH(Gr) in CH(Gr). We say that a cycle α
in CH(Gr) is rational if it belongs to CH(Gr). We use similar notations and definitions
for the cycles on Gr2 , the classes of cycles modulo 2, etc.
2
Corollary 87.3. The elements (ek × 1) + (1 × ek ) in CH(Gr ) are rational for all
k ∈ [1, n].
Proof. Let E1 and E2 be the two pull backs of E on Gr2 := Gr × Gr. Pulling the
2 2
formula 86.3 back to X × Gr , we get in CH(X × Gr ):
n
X
[P(E1 )] = ln−1 × 1 × 1 + hn−k × ek × 1,
k=1
n
X
[P(E2 )] = ln−1 × 1 × 1 + hn−k × 1 × ek .
k=1
Therefore the cycle
n
X
[P(E1 )] − [P(E2 )] = hn−k × (ek × 1 − 1 × ek )
k=1
is rational. Applying Proposition 87.1 to the variety Gr2 , we have the cycles (ek × 1) −
(1 × ek ) are also rational. Note that by Proposition 86.13, the cycles 2ek are rational. ¤
Now consider the Chow groups Ch(Gr), Ch(Gr) modulo 2 and write Ch(Gr) for the
image of Ch(Gr) in Ch(Gr). We still write ek for the class of the generator in Chk (Gr).
87. THE CHOW RING OF Gr(ϕ) IN THE GENERAL CASE 373
For every subset I ⊂ [1, n], the rational correspondence

Y£ ¤
(87.4) xI = (ek × 1) + (1 × ek ) ∈ Ch(Gr2 )
k∈I
defines an endomorphism (xI )∗ of Ch(Gr) taking Ch(Gr) into Ch(Gr).

Lemma 87.5. For any subsets I, J ⊂ [1, n], we have
½
eI∩J if I ∪ J = [1, n]
(xJ )∗ (eI ) =
0 otherwise
in Ch(Gr).
P
Proof. We have xJ = eJ1 × eJ2 , where the sum is taken over all subsets J1 and J2
of [1, n] such that J is the disjoint union of J1 and J2 . Hence
X
(xJ )∗ (eI ) = deg(eI · eJ1 ) eJ2
and the statement is implied by the following lemma. ¤ delete squar
Lemma 87.6. For any subsets I, J ⊂ [1, n],
½
1 mod 2 if J = [1, n] \ I
deg(eI · eJ ) ≡
0 mod 2 otherwise.
Proof. If J = [1, n] \ I, the product eI · eJ = e[1, n] is the class of a rational point of
Gr by Corollary 86.10, hence deg(eI · eJ ) = 1. Otherwise modulo 2, eI · eJ is either zero
or the monomial eK for some K different from [1, n] (one uses the relations between the
generators modulo 2). Hence deg(eI · eJ ) ≡ 0 mod 2. ¤
The following statement is due to Vishik. We give a proof different from the one in
[139].
Theorem 87.7. Let Gr be the variety of maximal isotropic subspaces of a non-degenerate
quadratic form of dimension 2n + 1 or 2n + 2. Then the ring Ch(Gr) is generated by all
ek , k ∈ [0, n], such that ek ∈ Ch(Gr).
Proof. By Propositions 85.2 and 86.17, it suffices to consider the case of dimension
2n + 1. It followsP from Theorem 86.12 that every element α ∈ Ch(Gr) can be written in
the form α = aI eI with aI ∈ Z/2Z. It suffices to prove the following:
Claim: For every I satisfying aI = 1, we have ek ∈ Ch(Gr) for any k ∈ I.
To proof the claim, we may assume that α is homogeneous. We establsh the claim by
induction on the number of nonzero coefficients of α. Choose I with largest |I| such that
aI = 1 and set J = ([1, n] \ I) ∪ {k}. By Lemma 87.5, (xJ )∗ (α) = ek or 1 + ek . Indeed,
if aI 0 = 1 for some I 0 ⊂ [1, n] with I 0 ∪ J = [1, n], then either I 0 = [1, n] \ J and hence
(xJ )∗ (eI 0 ) = e∅ = 1 or I 0 = ([1, n] \ J) ∪ {l} for some l. But since α is homogeneous, we
must have l = k. Therefore I 0 = I and (xJ )∗ (eI 0 ) = ek .
We have shown that ek ∈ Ch(Gr) for all k ∈ I. Therefore, eI ∈ Ch(Gr) and α − eI ∈
Ch(Gr). By the induction hypothesis, the claim holds for α − eI and therefore for α. ¤
V2 Exercise 87.8. Prove that the tangent bundle of Gr is canonically isomorphic to

(V /E).
88. The invariant J(ϕ)

In this section, we define a new invariant of non-degenerate quadratic forms. It differs
slightly from the one defined in [139].
Let ϕ be a non-degenerate quadratic form of dimension 2n + 1 or 2n + 2 and set
Gr = Gr(ϕ). As before let ek be viewed as the class of the generator for Ch(Gr). We
define a new discrete J-invariant J(ϕ) as follows:
/ Ch(Gr)}.
J(ϕ) = {k ∈ [0, n] with ek ∈
Recall that e0 = 1 if dim ϕ = 2n+1 hence in this case J(ϕ) ⊂ [1, n]. When dim ϕ = 2n+2,
we have 0 ∈ J(ϕ) if and only if the discriminant of ϕ is not trivial.
If dim ϕ = 2n + 2 and ϕ0 is a non-degenerate subform of ϕ of codimension 1, then
½
J(ϕ0 ) if disc ϕ is trivial
J(ϕ) =
{0} ∪ J(ϕ0 ) otherwise.
For a subset I ⊂ [0, n] let ||I|| denote the sum of all k ∈ I.
Proposition 88.1. The smallest dimension i such that Chi (Gr) 6= 0 is equal to
||J(ϕ)||.
Proof. By Theorem 87.7, the product of all ek satisfying k ∈/ J(ϕ) is a nontrivial
element of Ch(Gr) of the smallest dimension equal to ||J(ϕ)||. ¤
Proposition 88.2. A non-degenerate quadratic form ϕ is split if and only if J(ϕ) = ∅.
Proof. The “only if” part follows from the definition. Suppose the set J(ϕ) is empty.
Since all the ek are rational, the class of a rational point of Gr belongs to Ch0 (Gr) by
Corollary 86.10. It follows that Gr has a closed point of odd degree, i.e., ϕ is split over
an odd degree finite field extension. By Springer’s theorem (Corollary 18.5), the form ϕ
is split. ¤
Lemma 88.3. Let ϕ = ϕ̃ ⊥ H. Then J(ϕ) = J(ϕ̃).
Proof. Suppose that dim ϕ = 2n + 1. Note first that the cycle en = [Gr(ϕ̃)] is
rational so that n ∈/ J(ϕ). Let k ≤ n − 1. It follows from the decomposition (86.2) that
CHk (Gr) ' CHk Gr(ϕ̃) and ek corresponds to ẽk by Lemma 86.11. Hence ek ∈ J(ϕ) if
and only if ẽk ∈ J(ϕ̃). The case of the even dimension is similar. ¤
Corollary 88.4. Let ϕ and ϕ0 be Witt-equivalent quadratic forms. Then J(ϕ) =
J(ϕ0 ).
Lemma 88.5. Let X be a variety,
¡ Y ¢a scheme and n an integer such that the natural
homomorphism CHi (X) → CHi ¡XF (y) ¢ is surjective for every point y ∈ Y and i ≥
dim X − n. Then CHj (Y ) → CHj YF (X) is surjective for every j ≥ dim Y − n.
Proof. Using a localization argument similar to that used in the proof Proposition
52.10, one checks that the top homomorphism in the commutative diagram
CH(X) ⊗ CH(Y ) −−−→ CH(X × Y )
 
 
y y
¡ ¢
CH(Y ) −−−→ CH YF (X)
88. THE INVARIANT J(ϕ) 375
is surjective in dimensions ≥ dim X + dim Y − n by induction on dim Y . Since the

right vertical homomorphism is surjective, so is the bottom homomorphism in dimensions
≥ dim Y − n. ¤
Let ϕ be a quadratic form of dimension 2n + 1 or 2n + 2.
¡ ¢
Corollary 88.6. The canonical homomorphism CHi (Gr) → CHi GrF (X) is surjec-
tive for all i ≤ n − 1.
Proof. Note that X is split over F (y) for every y ∈ Gr. Therefore, CHk (X¡F (y) ) is¢
generated by hk for all k ≤ n − 1 and hence the homomorphism CHk (X) → CHk XF (y)
is surjective. ¤
Corollary 88.7. J(ϕ) ∩ [0, n − 1] ⊂ J(ϕF (X) ) ⊂ J(ϕ).
The following proposition relates the set J(ϕ) and the absolute Witt indices of ϕ. It
follows from Corollaries 88.4 and 88.7.
Proposition 88.8. Let ϕ be a non-degenerate quadratic form of dimension 2n + 1 or
2n + 2. Then © ª
J(ϕ) ⊂ n − j0 (ϕ), n − j1 (ϕ), . . . , n − jh(ϕ)−1 (ϕ) .
In particular, |J(ϕ)| ≤ h(ϕ).
Remark 88.9. One can impose further restrictions on J(ϕ). Choose a non-degenerate
form ψ such that one of the forms ϕ and ψ is a subform of the other of codimension 1 and
the dimension of the largest form is even. Then the sets J(ϕ) and J(ψ) differ by at most
one element 0. Therefore, the inclusion in Proposition 88.8 applied to the form ψ gives
© ª
J(ϕ) ⊂ 0, n − j0 (ψ), n − j1 (ψ), . . . , n − jh(ψ)−1 (ψ) .
Example 88.10. Suppose that ϕ is an anisotropic m-fold Pfister form, m ≥ 1. Then
J(ϕ) = {2m−1 − 1}. Indeed, h(ϕ) = 1 hence J(ϕ) ⊂ {2m−1 − 1} by Proposition 88.8. But
J(ϕ) is not empty by Proposition 88.2.
We write nGr for the gcd of deg(g) taken over all closed points g ∈ Gr. The ideal
nGr · Z is the image of the degree homomorphism CH(Gr) → Z. Since ϕ splits over a field
extension of F of degree a power of 2, the number nGr is a 2-power.
Proposition 88.11. Let ϕ be a non-degenerate quadratic form of odd dimension.
Then ¡ ¢
2|J(ϕ)| · Z ⊂ nGr · Z ⊂ ind C0 (ϕ) · Z.
k
Proof. For every k ∈ / J(ϕ), let fk be a cycle in CH (Gr) satisfying fk ≡ ek modulo
k k
2 CH (Gr). By Remark 86.14, we have 2ek ∈ CH (Gr) for all k. Let α be the product of
all the fk satisfying k ∈
/ J(ϕ) and all the 2ek satisfying k ∈ J(ϕ). Clearly, α is a cycle
|J(ϕ)|
in CH(Gr) of degree 2 m, where m is an odd integer. The first inclusion now follows
from the fact that nGr is a 2-power.
Let L be the residue field F (g) of a closed point g ∈ Gr. Since ϕ splits over L, so does
the even Clifford algebra C0 (ϕ). It follows that ind C0 (ϕ) divides [L : F ] = deg g for all g
and therefore divides nGr . ¤
Propositions 88.8 and 88.11 yield

Corollary 88.12. Let ϕ be a non-degenerate quadratic form of dimension 2n + 1.
Consider the following statements:
(1) C0 (ϕ) is a division algebra.
(2) nGr = 2n .
(3) J(ϕ) = [1, n].
(4) jk = k for all k = 0, 1, . . . , n.
Then (1) ⇒ (2) ⇒ (3) ⇒ (4).
The following statement is a refinement of the implication (1) ⇒ (3).
Corollary 88.13. Let ϕ be a non-degenerate quadratic form of odd dimension and
ind C0 (ϕ) = 2k . Then [1, k] ⊂ J(ϕ).
Proof. We proceed by induction on dim ϕ = 2n + 1. If k = n, i.e., C0 (ϕ) is a
division algebra, the statement follows from Corollary 88.12. So we may assume that
k < n. Let ϕ0 be a form over F (ϕ) Witt-equivalent to ϕF (ϕ) of dimension less than dim ϕ.
The even Clifford algebra C0 (ϕ0 ) is Brauer-equivalent to C¡0 (ϕ)F (ϕ)¢. Since¡ C0 (ϕ)¢ is not
a division algebra, it follows from Corollary 30.10 that ind C0 (ϕ0 ) = ind C0 (ϕ) = 2k .
By the induction hypothesis, [1, k] ⊂ J(ϕ0 ). By Corollaries 88.4 and 88.7, we have
J(ϕ0 ) = J(ϕF (ϕ) ) ⊂ J(ϕ). ¤
Exercise 88.14. Let ϕ be a quadratic form of odd dimension.
(1) Prove that 1 ∈ J(ϕ) if and only if the even Clifford algebra C0 (ϕ) is not split.
(2) Prove that 2 ∈ J(ϕ) if and only if ind C0 (ϕ) > 2.
¡ ¢
89. Steenrod operations on Ch Gr(ϕ)
¡ ¢
We calculate the Steenrod operations on Ch Gr(ϕ) as given by Vishik in [139].
Let ϕ be a non-degenerate quadratic form on V over F of dimension 2n + 1 or 2n + 2,
X the projective quadric of ϕ, Gr the variety of maximal totally isotropic subspaces of
V , and E the tautological vector bundle over Gr. Let s : P(E) → Gr and t : P(E) → X
be the projections. There is an exact sequence of vector bundles over P(E)
0 → 1 → L⊕n
c → Tt → 0,
where Lc is the canonical line bundle over P(E) and Tt is the relative tangent bundle of t
(cf. Example 104.20). Note that Lc is the pull-back with respect to t of the canonical line
bundle over X, hence c(Lc ) = 1 + t∗ (h),
¡ where ¢ h ∈ CH1 (X) is the class of a hyperplane
n
section of X. It follows that c(Tt ) = 1 + t∗ (h) .
Theorem 89.1. Let char F 6= 2 and Gr = Gr(ϕ) with ϕ a non-degenerate quadratic
form of dimension 2n+1 or 2n+2. Then the Steenrod operation SqGr : Ch(Gr) → Ch(Gr)
of cohomological type satisfies µ ¶
i k
SqGr (ek ) = ek+i
i
for all i and k ∈ [1, n].
90. CANONICAL DIMENSION 377
Proof. We have SqX (ln−k ) = (1 + h)n+k · ln−k by Corollary 78.5. It follows from
(86.5), Theorem 61.8, and Proposition 61.9 that
SqGr (ek ) = SqGr ◦s∗ ◦ t∗ (ln−k )
= s∗ ◦ c(−Tt ) ◦ SqP(E) ◦t∗ (ln−k )
¡ ¢
= s∗ (1 + t∗ h)−n · t∗ ◦ SqX (ln−k )
¡ ¢
= s∗ ◦ t∗ (1 + h)−n · (1 + h)n+k · ln−k
¡ ¢
= s∗ ◦ t∗ (1 + h)k · ln−k
X µk ¶
= s∗ ◦ t∗ (ln−k−i )
i≥0
i
X k¶ µ
= ek+i . ¤
i≥0
i
Exercise 89.2. Let ϕ be an anisotropic quadratic form of even dimension and height
1. Using Steenrod operations, give another proof of the fact that dim(ϕ) is a 2-power.
(Hint: Use Propositions 88.2 and 88.8.)
90. Canonical dimension
Let F be a field and let C be a class of field extensions of F . Call a field E ∈ C generic
if for any L ∈ C there is an F -place E * L (cf. §103).
Example 90.1. Let X be a scheme over F . A field extensions L of F is called an
isotropy field of X if X(L) 6= ∅. If X is a smooth variety, it follows from §103 that the
field F (X) is generic in the class of all isotropy fields of X.
The canonical dimension cdim(C) of the class C is defined to be the minimum of the
tr. degF E over all generic fields E ∈ C. If X is a scheme over F , we write cdim(X) for
cdim(C), where C is the class of fields as defined in Example 90.1. If X is smooth then
cdim(X) ≤ dim X.
Let p be a prime integer and C a class of field extensions of F . A field E ∈ C is called
p-generic if for any L ∈ C there is an F -place E * L0 for some finite extension L0 of L of
degree prime to p. The canonical p-dimension cdimp (C) of C and cdimp (X) of a scheme
X over F are defined similarly. Clearly, cdimp (C) ≤ cdim(C) and cdimp (X) ≤ cdim(X).
The following theorem answers an old question of Knebusch (cf. [80, §4]):
Theorem 90.2. For an arbitrary anisotropic smooth projective quadric X,
cdim2 (X) = cdim(X) = dimIzh X.
Proof. Let Y be a smooth subquadric of X of dimension dim Y = dimIzh X. Note
that i1 (Y ) = 1 by Corollary 74.3. Clearly, the function field F (Y ) is an isotropy field of
X. Moreover, if L is an isotropy field of X, then by Lemma 74.1, we have Y (L) 6= ∅.
Since the variety Y is smooth, there is an F -place F (Y ) * L (cf. §103). Therefore, F (Y )
is a generic isotropy field of X.
Suppose that E is an arbitrary 2-generic isotropy field of X. We show that tr. degF E ≥
dim Y which will finish the proof.
Since E and F (Y ) are both generic isotropy fields of the same X, we have F -places
π : F (Y ) * E and ε : E * E 0 , where E 0 is an odd degree field extension of F (Y ). Let y
and y 0 be the centers of π and ε ◦ π respectively. Clearly, y 0 is a specialization of y and
therefore,
dim y 0 ≤ dim y ≤ tr. degF E.
The morphism Spec E 0 → Y induced by ε ◦ π gives rise to a prime correspondence δ :
Y Ã Y with odd mult(δ) so that p2∗ (δ) = [y 0 ], where p2 : Y × Y → Y is the second
projection. By Theorem 75.4, mult(δ t ) is odd, hence y 0 is the generic point of Y and
dim y 0 = dim Y . ¤
The rest of this section is dedicated to determining the canonical 2-dimension of the
class C of all splitting fields of a non-degenerate quadratic form ϕ. Note for this class C,
we have cdim(C) = cdim(Gr) and cdim2 (C) = cdim2 (Gr), where Gr = Gr(ϕ), since L ∈ C
if and only if Gr(L) 6= ∅. Hence
cdim2 (Gr) ≤ cdim(Gr) ≤ dim Gr .
Theorem 90.3. Let ϕ be a non-degenerate quadratic form over F . Then

¡ ¢
cdim2 Gr(ϕ) = ||J(ϕ)||.
Proof. Let E be a 2-generic isotropy field of Gr such that tr. degF E = cdim(Gr). As
E is an isotropy field, there is a morphism Spec E → Gr over F . Let Y be the closure of the
image of this morphism. We view F (Y ) as a subfield of E. Clearly, tr. degF E ≥ dim Y .
Since E is 2-generic, there is a field extension L/F (Gr) of odd degree and an F -place
E * L. Restricting this place to the subfield F (Y ), we get a morphism f : Spec L → Y
as Y is complete. Let g : Spec L → Gr be the morphism induced by the field extension
L/F (Gr). Then the closure Z of the image of the morphism (f, g) : Spec L → Y × Gr is
of odd degree [L : F (Gr)] when projecting to Gr. Therefore, the image of [Z] under the
composition
(i×1Gr )∗ q∗
Ch(Y × Gr) −−−−−→ Ch(Gr × Gr) −
→ Ch(Gr),
where i : Y → Gr is the closed embedding and q is the second projection, is equal to [Gr].
In particular, (i × 1Gr )∗ ([Z]) 6= 0, hence (i × 1Gr )∗ 6= 0.
We claim that the push-forward homomorphism i∗ : Ch(Y ) → CH(Gr) is also non-
trivial. Let L be the residue field of a point of Y . Consider the induced morphism
j : Spec L → Gr. The pull-back of the element xI in Ch(Gr2 ) (defined in (87.4)) with
respect to the morphism j × 1Gr : GrL → Gr2 is equal to eI ∈ Ch(GrL ) = Ch(GrL ).
Since the elements eI generate Ch(GrL ) by Theorem 86.12, the pull-back homomor-
phism Ch(Gr2 ) → Ch(GrL ) is surjective. Applying Proposition 58.18 to the projection
p : Y × Gr → Y and the embedding i × 1Gr : Y × Gr → Gr2 , we conclude that the product
hY : Ch(Y ) ⊗ Ch(Gr2 ) → Ch(Y × Gr), α ⊗ β 7→ p∗ (α) · β
is surjective.
90. CANONICAL DIMENSION 379
By Proposition 58.17, the diagram

h
Ch(Y ) ⊗ Ch(Gr2 ) −−−
Y
→ Ch(Y × Gr)
 

i∗ ⊗1y
(i×1 )
y Gr ∗
h
Ch(Gr) ⊗ Ch(Gr2 ) −−−
Gr
→ Ch(Gr × Gr)
is commutative. As (i × 1Gr )∗ is nontrivial, we conclude that i∗ is also nontrivial. This
proves the claim.
By Proposition 88.1, we have dim Y ≥ ||J(ϕ)||, hence
cdim2 (Gr) = tr. degF E ≥ dim Y ≥ ||J(ϕ)||.
It follows from Proposition 88.1 that there is a closed subvariety Y ⊂ Gr of dimension
||J(ϕ)|| such that [Y ] 6= 0 in Ch(Gr) = Ch(GrF (Gr) ). By Lemma 87.6, there is a β ∈
Ch(GrF (Gr) ) satisfying [Y ] · β 6= 0 in Ch(GrF (Gr) ). It follows from Proposition 56.11 that
the product [Y ] · β belongs to the image of the push-forward homomorphism
¡ ¢ ¡ ¢
Ch YF (Gr) → Ch GrF (Gr) ,
¡ ¢
therefore Ch0 YF (Gr) 6= 0. In other words, there is a closed point y ∈ YF (Gr) of odd degree.
Let Z be the closure of the image of y under the canonical morphism YF (Gr) → Y × Gr.
Note that that the projection Z → Gr is of odd degree deg(y), hence F (Z) is an extension
of F (Gr) of odd degree. Let Y 0 denote the image of the projection Z → Y , so F (Y 0 ) is
isomorphic to a subfield of F (Z).
We claim that F (Y 0 ) is a 2-generic splitting field of Gr. Indeed, since Y 0 is a subvariety
of Gr, the field F (Y 0 ) is a splitting field of Gr. Let L be another splitting field of Gr. By
Lemma 103.2, a geometric F -place F (Gr) * L can be extended to an F -place F (Z) * L0
where L0 is an extension of L of odd degree. Restricting to F (Y 0 ), we get an F -place
F (Y 0 ) * L0 . This proves the claim. Therefore, we have
cdim2 (Gr) ≤ dim Y 0 ≤ dim Y = ||J(ϕ)||. ¤
Theorem 90.3 and Corollary 88.12 yield
Corollary 90.4. Let ϕ be a non-degenerate quadratic form of dimension 2n + 1 such
that J(ϕ) = [1, n] (e.g., if C0 (ϕ) is a division algebra or if nGr = 2n ). Then
n(n + 1)
cdim2 (Gr) = cdim(Gr) = dim(Gr) = .
2
Example 90.5. Let ϕ be an anisotropic m-fold Pfister form with m ≥ 1. Since the
class of splitting fields of ϕ coincides with the class of isotropy fields, we have cdim(Gr) =
dimIzh (X) = 2m−1 − 1. By Theorem 90.3 and Example 88.10, we have cdim2 (Gr) =
||J(ϕ)|| = 2m−1 − 1.
We next compute the canonical dimensions cdim(Gr), cdim2 (Gr) and determine the
set J(ϕ) for an excellent quadratic form ϕ. Write the dimension of ϕ in the form
(90.6) dim ϕ = 2p0 − 2p1 + 2p2 − · · · + (−1)r−1 2pr−1 + (−1)r 2pr
with some integers p0 , p1 , . . . , pr satisfying p0 > p1 > · · · > pr−1 > pr + 1 > 0. Note that
the height h of ϕ equals r + 1 for even dim ϕ, while h = r if dim ϕ is odd.
Let ψ be the leading ph -fold Pfister of ϕ (defined over F ). Since ϕ and ψ have the
same classes of splitting fields, we have cdim Gr(ϕ) = cdim Gr(ψ) and cdim2 Gr(ϕ) =
cdim2 Gr(ψ). By Example 90.5,
¡ ¢ ¡ ¢
(90.7) cdim Gr(ϕ) = cdim2 Gr(ϕ) = 2ph −1 − 1.
Proposition 90.8. Let ϕ be an anisotropic excellent form of height h. Then J(ϕ) =
{2ph −1 − 1}, where the integer ph is determined in (90.6).
Proof. Note that jh−1 = (dim ϕ − dim ψ)/2 where ψ is as above. Hence by Propo-
sition ¡88.8, every
¢ element of J(ϕ) is at least 2ph −1 − 1. By Theorem 90.3, we have
cdim2 Gr(ϕ) = ||J(ϕ)||. It follows from (90.7) that J(ϕ) = {2ph −1 − 1}. ¤
The notion of canonical dimension of algebraic groups was introduced by Berhuy and
Reichstein in [18]. Here we have presented the more general notion of canonical dimension
and p-canonical dimension
¡ of¢ a class of field extensions of a given field (cf. [75]). The
computation of cdim¡2 Gr(ϕ) ¢ given in ¡ Theorem¢ 90.3 is new. It is conjectured in [139,
Conj. 6.6] that cdim Gr(ϕ) = cdim2 Gr(ϕ) .
CHAPTER XVII
Motives of quadrics
91. Comparison of some discrete invariants of quadratic forms

In this section, F is an arbitrary field, n a positive integer, V a vector space over F
of dimension 2n or 2n + 1, ϕ a non-degenerate quadratic form on V , X the projective
quadric of ϕ. For any positive integer i, we write Gi for the scheme of i-dimensional
totally isotropic subspaces of V . In particular, G1 = X and Gi = ∅ for i > n.
We write Ch(Y ) for the Chow group modulo 2 of an F -scheme Y ; Ch(Ȳ ) is the colimit
of Ch(YL ) over all field extensions L/F , Ch(Y ) is the reduced Chow group, i.e., the image
of the homomorphism Ch(Y ) → Ch(Ȳ ). L
We write Ch(G∗ ) for the direct sum i≥1 Ch(Gi ). We recall that Ch(X ∗ ) stands for
L i i
i≥1 Ch(X ), where X is the direct product of i copies of X. We consider Ch(G∗ ) and
Ch(X ∗ ) as invariants of the quadratic form ϕ. Note that their components Ch(Gi ) and
Ch(X i ) are subsets of the finite sets Ch(Ḡi ) and Ch(X̄ i ) depending only on dim ϕ.
These invariants are not independent. A relation between them is described in the
following theorem:
Theorem 91.1. The following three invariants of a non-degenerate quadratic form
ϕ of a fixed dimension are equivalent in the sense that if ϕ0 is another non-degenerate
quadratic form with dim ϕ = dim ϕ0 and the values of one of the invariants for ϕ and ϕ0
are equal then the values of any other of the invariants for ϕ and ϕ0 are also equal.
(i) Ch(X ∗ )
(ii) Ch(X n )
(iii) Ch(G∗ ).
Here X = Xϕ and G = G(ϕ).
Remark 91.2. Although the equivalence of the above invariants means that any of
them can be expressed in terms of any other, it does not seem to be possible to get
manageable formulas relating (iii) with (ii) or (i).
We need some preparation to prove Theorem 91.1, . For i ≥ 1, let us write Fli for
the scheme of flags V1 ⊂ · · · ⊂ Vi of totally isotropic subspaces V1 , . . . , Vi of V , where
dim Vj = j. In particular, Fl1 = X and Fli = ∅ for i > n. The following lemma generalizes
Example 66.6:
Lemma 91.3. For any i ≥ 1, the product Fli ×X has the canonical structure of a
relative cellular scheme with a basis of cells given by
0) a projective bundle over Fli ,
1) the scheme Fli+1 ,
381
382 XVII. MOTIVES OF QUADRICS
2) the scheme Fli .

Proof. We construct the cellular filtration
Y0 ⊂ Y1 ⊂ Y2 = Y
on the scheme Y = Fli ×X as follows: Y1 is the subscheme of pairs
(V1 ⊂ · · · ⊂ Vi , W )
such that the subspace W + Vi is totally isotropic and Y0 is the subscheme of pairs with
W ⊂ Vi . The projection of the scheme Y0 onto Fli is a (rank i − 1) projective bundle.
Of course, if i ≥ n then Y0 = Y1 (and the base of the “cell” Y1 \ Y0 is the empty scheme
Fli+1 ). ¤
Corollary 91.4. The motive of the product Fli ×X canonically decomposes as a
direct sum, where each summand is some shift of the motive of the scheme Fli or of the
scheme Fli+1 . Moreover, a shift of the motive of Fli occurs for every i ≤ n and a shift of
the motive of Fli+1 also occurs for every i + 1 ≤ n.
Proof. By Corollary 66.4 and Lemma 91.3, the motive of Fli ×X decomposes into a
direct sum of three summands that are some shifts of the motives of Y0 , Fli+1 , and Fli ,
where Y0 is a projective bundle over Fli . By Theorem 63.10, the motive of Y0 is also a
direct sum of shifts of the motive of Fli . ¤
Corollary 91.5. For any r ≥ 1, the motive of X r canonically decomposes into a
direct sum, where each summand is a shift of the motive of some Fli with i ∈ [1, r].
Moreover, for any i ∈ [1, r] with i ≤ n, a shift of the motive of Fli occurs.
Proof. We induct on r. Since X 1 = X = Fl1 , the case r = 1 is immediate. If the
statement is proved for some r ≥ 1, then the statement for X r+1 follows by Corollary
91.4. ¤
Lemma 91.6. For any i ≥ 1, the motive of Fli canonically decomposes into a direct
sum, where each summand is a shift of the motive of the scheme Gi .
Proof. For each j ∈ [1, i], write Φj for the scheme of flags V1 ⊂ · · · ⊂ Vi−j ⊂ Vi of
totally isotropic subspaces Vk of V satisfying dim Vk = k for any k. In particular, Φ1 = Fli
and Φi = Gi . The projections
Fli = Φ1 → Φ2 → · · · → Φi = Gi
are projective bundles. Therefore, the lemma follows from Theorem 63.10. ¤
Combining Corollary 91.5 with Lemma 91.6, we get
Corollary 91.7. For any r ≥ 1, the motive of X r canonically decomposes into a
direct sum, where each summand is a shift of the motive of some Gi with i ∈ [1, r].
Moreover, for any i ∈ [1, r] with i ≤ n, a shift of the motive of Gi occurs.
Proof of Theorem 91.1. The equivalences (i) ⇔ (iii) and (ii) ⇔ (iii) are given
by Corollary 91.7. ¤
92. THE NILPOTENCE THEOREM FOR QUADRICS 383
Remark 91.8. One may say that the invariant Ch(X n ) is a “compact form” of the
invariant Ch(X ∗ ) and also that the invariant Ch(G∗ ) is a “compact form” of Ch(X n ).
However, some properties of these invariants are easier formulated and proven on the
level of Ch(X ∗ ). Among such properties (used many times above), we have the stability
of Ch(X ∗ ) ⊂ Ch(X̄ ∗ ) under partial operations on Ch(X̄ ∗ ) given by permutations of factors
of any X r as well as pull-backs and push-forwards with respect to partial projections and
partial diagonals between X r and X r+1 . It is also easier to describe a basis of Ch(X̄ ∗ ) and
compute multiplication and Steenrod operations (giving further restrictions on Ch(X ∗ ))
in terms of this basis, than to do a similar analysis for Ch(Ḡ∗ ).
92. The Nilpotence Theorem for quadrics

Let Λ be a commutative ring. We shall work in the categories CR∗ (F, Λ) and CR(F, Λ),
introduced in §63.
Let C be a class of smooth complete schemes over field extensions of F closed under tak-
ing finite disjoint unions (of schemes over the same field), taking connected components,
and scalar extensions. We say that C is tractable if for any variety X in C having a rational
point and of positive dimension, there is a scheme X 0 in C satisfying dim X 0 < dim X and
M (X 0 ) ' M (X) in CR∗ (F, Λ). A scheme is called tractable, if it is member of a tractable
class.
Our primary example of a tractable scheme will be any smooth projective quadric over
F , the tractable class being the class of (all finite disjoint unions) of all smooth projective
quadrics over field extensions of F (cf. Example 66.7).
A smooth projective scheme is called split if its motive in CR∗ (F, Λ) is isomorphic to
a finite direct sum of several copies of the motive Λ. Any tractable scheme X splits over
an extension of the base field. Moreover, the number of copies of Λ in the corresponding
decomposition is an invariant of X. We call this the rank of X and denote it by rk X.
The number of components of any tractable scheme does not exceed its rank.
Exercise 92.1. Let X/F be a smooth complete variety such that for any field exten-
sion E/F satisfying X(E) 6= ∅, the scheme XE is split (for example, the variety of the
maximal totally isotropic subspace of a non-degenerate odd-dimensional quadratic form
considered in Chapter XVI). Show that X is tractable.
Exercise 92.2. Show that the product of two tractable schemes is tractable.
Remark 92.3. As shown in [27], the class of all projective homogenous varieties
(under the action of an algebraic group) is tractable.
The following theorem was initially proved by Rost in the case of quadrics. The more
general case of a projective homogeneous variety was done in [27].
Theorem 92.4 (Nilpotence Theorem for tractable schemes). Let X be a tractable
scheme over F with M (X) its motive in CR∗ (F, Λ) or in CR(F, Λ) and α ∈ End M (X)
a correspondence. If αE ∈ End M (XE ) vanishes for some field extension E/F then α is
nilpotent.
Proof. It suffices to consider the case of the category CR∗ (F, Λ) as the functor of
(63.3) is faithful. We fix a tractable class of schemes containing X. We shall construct a
map
N : [0, +∞) × [1, rk X] → [1, +∞)
with the following properties. If Y is a scheme in the tractable class with rk Y ≤ rk X and
α ∈ CH(Y 2 ; Λ) a correspondence vanishing over some field extension of F then αN (i,j) = 0
if dim Y ≤ i and the number of i-dimensional connected components of Y is at most j.
If dim Y = 0 then any extension of scalars induces an injection of CH(Y 2 ; Λ). In this
case, we set N (0, i) = 1 for any i ≥ 1.
Now order the set [0, +∞) × [1, rk X] lexicographically. Let (i, j) be a pair with
i ≥ 1. Assume that N has been defined on all pairs smaller than (i, j).
Let Y be an arbitrary scheme in the class such that dim Y = i and the number
of the i-dimensional components of Y is j. To simplify the notation, we shall assume
that the field of definition of Y is F . Let Y1 be a fixed i-dimensional component of
Y and set Y0 to be the union of the remaining components of Y . We take an arbitrary
0 0
correspondence α ∈ CH(Y 2 ; Λ) vanishing over a scalar extension and replace it by αN (i ,j ) ,
where (i0 , j 0 ) < (i, j) is the lexicographical order. Then for any point y ∈ Y1 , we have
αF (y) = 0, since the motive of the scheme YF (y) is isomorphic to the motive of another
scheme having j − 1 components of dimension i. Applying Theorem 67.1, we see that
αi+1 ◦ CH(Y1 × Y ; Λ) = 0 .
In particular, the composite of the inclusion morphism M (Y1 ) → M (Y ) with αi+1 is
trivial. Replace α by L αi+1 . We can view α as a 2 × 2 matrix according
¡ to the decomposi-
¢
tion M (Y ) ' M (Y0 ) M (Y1 ). Its entries corresponding to Hom M (Y1 ), M (Y0 ) and to
End M (Y1 ) are 0. Moreover, the matrix entry corresponding to End M (Y0 ) is nilpotent
with N (i0 , j 0 ) its nilpotence exponent, as the number of the i-dimensional components of
0 0
Y0 is at most j − 1. Replacing α by αN (i ,j ) , we may assume that α has only¡ one possibly
¢
nonzero entry, namely, the (non-diagonal) entry corresponding to Hom M (Y0 ), M (Y1 ) .
Therefore, α2 = 0. Set N (i, j) = 2(i + 1)N (i0 , j 0 )2 . We have shown that for any scheme
Y in the tractable class with rk Y ≤ rk X and any correspondence α ∈ CH(Y 2 ; Λ) van-
ishing over some field extension of F we have αN (i,j) = 0 if dim Y = i and the number
of i-dimensional connected components of Y is j. Since N (i, j) ≥ N (i0 , j 0 ), one also has
αN (i,j) = 0 if dim Y ≤ i and the number of i-dimensional connected components of Y is
smaller than j. ¤
Corollary 92.5. Let X be a tractable scheme over F and E/F a field extension.
If q ∈ End M (XE ) is an idempotent lying in the image of the restriction End M (X) →
End M (XE ) in the motivic category CR∗ (F, Λ) or CR(F, Λ) then there exists an idempo-
tent p ∈ End M (X) satisfying pE = q.
Proof. Choose a correspondence p0 ∈ End M (X) satisfying p0E = q. Let ¡ A (re-¢
spectively, B) be the (commutative) subring of End M (X) (respectively, End M (XE ) )
generated by p0 (respectively, q). By Theorem 92.4, the kernel of the ring epimorphism
A ³ B consists of nilpotent elements. It follows that the map Spec B → Spec A is a
homeomorphism and, in particular, induces a bijection of the sets of the connected com-
ponents of these topological spaces. Therefore, the homomorphism A → B induces a
93. CRITERION OF ISOMORPHISM 385
bijection of the sets of idempotents of these rings (cf. [19, Cor. 1 to Prop. 15 of §4.3 of
Ch. II]), and we can find a required p inside of A. ¤
Exercise 92.6. Show that one can take for p some power of p0 . Hint: prove and use
the fact that the kernel of End X → End XE is annihilated by some positive integer.
Corollary 92.7. Let X and Y be tractable schemes and p ∈ End M (X) and q ∈
End M (Y ) idempotents in the motivic category CR∗ (F, Λ) or CR(F, Λ). Let f be a mor-
phism (X, p) → (Y, q) in the category CM∗ (F, Λ) or CM(F, Λ) respectively. If fE is an
isomorphism for some field extension E/F then f is also an isomorphism.
Proof. By Proposition 63.4, it suffices to prove the result for the category CR∗ (F, Λ).
First suppose that Y = X and q = p. We may assume that the scheme XE is split and
we have fixed an isomorphism of the motive (X, p) with the direct sum of n copies of Λ for
some n. Then Aut(XE , pE ) = GLn (Λ). Let P (t) ∈ Λ[t] be the characteristic polynomial
of the matrix fE , hence P (fE ) = 0. If Q(t) ∈ Λ[t] satisfies P (t) = P (0) + tQ(t), the
endomorphism
fE ◦ Q(fE ) = Q(fE ) ◦ fE = P (fE ) − P (0) = −P (0) = ± det fE
is multiplication by an invertible element ε = ± det fE in the coefficient ring Λ. By
Theorem 92.4, the endomorphisms α, β ∈ End(X, p) satisfying f ◦Q(f ) = ε+α and Q(f )◦
f = ε + β are nilpotent. Thus the composites f ◦ Q(f ) and Q(f ) ◦ f are automorphisms.
Consequently f is an automorphism.
In the general case, consider the transpose f t : (Y, q) → (X, p) of f . Since fE is
an isomorphism, fEt is also an isomorphism. It follows by the previous case that the
composites f ◦ f t and f t ◦ f are automorphisms. Thus f is an isomorphism. ¤
Corollary 92.8. Let X be a tractable scheme with p, p0 ∈ End M (X) idempotents
satisfying pE = p0E for some field E ⊃ F . Then the motives (X, p) and (X, p0 ) are
canonically isomorphic.
Proof. The morphism p0 ◦ p : (X, p) → (X, p0 ) is an isomorphism because it becomes
an isomorphism over E. ¤
93. Criterion of isomorphism
In this section, we let Λ = Z/2Z.
Theorem 93.1. Let X and Y be smooth projective quadrics over F . Then the motives
of X and Y in the category CR(F, Z/2Z) are isomorphic if and only if dim X = dim Y
and i0 (XL ) = i0 (YL ) for any field extension L/F .
Proof. The “only if” part of the statement is easy: the motive M (X) of X in
CR(F, Z/2Z) determines the graded group Ch∗ (X) which in turn determines dim X and
i0 (X) by Corollary 72.6. To prove the “if” part assume that dim X = dim Y and i0 (XL ) =
i0 (YL ) for any field extension L/F . As before we write D for dim X and set d = [D/2].
The case of split X and Y is trivial. Note that in the split case an isomorphism
M (X) → M (Y ) is given by the cycle cXY + deg(ld2 )(hd × hd ), where
P
d
cXY = (hi × li + li × hi ) ∈ Ch(X × Y )
i=0
(cf. Lemma 73.1). By Corollary 92.7, it follows in the non-split case that the motives of
X and Y are isomorphic if the cycle cXY ∈ Ch(X̄ × Ȳ ) is rational.
To prove Theorem 93.1 in the general case, we show by induction on D that the cycle
cXY is rational.
If X (and therefore Y ) is isotropic, then the cycle cX0 Y0 is rational by the induction
hypothesis , where X0 and Y0 are the anisotropic parts of X and Y respectively. It follows
that the cycle cXY is rational in the isotropic case. We may therefore assume that X and
Y are anisotropic.
To finish the proof we need two results. For their proofs we introduce some special
notation and terminology. Write N for the set of the symbols {hi × li , li × hi }i∈[0, d] .
For any subset I ⊂ N , write cXY (I) for the sum of the basis elements of ChD (X̄ × Ȳ )
corresponding to the symbols of I. Similarly, define the cycles cY X (I) ∈ ChD (Ȳ × X̄),
cXX (I) ∈ ChD (X̄ 2 ), and cY Y (I) ∈ ChD (Ȳ 2 ).
We call a subset I ⊂ N admissible, if the cycles cXY (I) and cY X (I) are rational and
weakly admissible if cXX (I) and cY Y (I) are rational.
Since the set N is weakly admissible by Lemma 73.1, the complement N \ I of any
weakly admissible set I is also weakly admissible.
Call a subset I ⊂ N symmetric, if it is stable under transposition, i.e., I t = I. For
any I ⊂ N , the set I ∪ I t is the smallest symmetric set containing I; we call it the
symmetrization of I.
Lemma 93.2. (1) Any admissible set is weakly admissible.
(2) The symmetrization of an admissible set is admissible.
(3) A union of admissible sets is admissible.
Proof. (1). This follows from the formulas which hold up to the addition of hd × hd :
cXX (I) = cY X (I) ◦ cXY (I) and cY Y (I) = cXY (I) ◦ cY X (I) .
(3). Let I and J be admissible sets. The cycle cXY (I ∪ J) is rational as
cXY (I ∪ J) = cXY (I) + cXY (J) + cXY (I ∩ J)
and up to the addition of hd × hd , we have cXY (I ∩ J) = cXY (J) ◦ cXX (I). The rationality
of cY X (I ∪ J) is proved analogously.
(2). The transpose I t of an admissible set I ⊂ N is admissible. Therefore, by (3), the
union I ∪ I t is admissible. ¤
The key observation is:
Proposition 93.3. Let I be a weakly admissible set and hr × lr ∈ I the element with
smallest r. Then hr × lr is contained in an admissible set.
Assuming Proposition 93.3, we finish the proof of Theorem 93.1 by showing that the
set N is admissible.
Note that ∅ is a symmetric admissible set. Let I0 be a symmetric admissible set.
It suffices to show that if I0 6= N then I0 is contained in a strictly bigger symmetric
admissible set I1 .
By Lemma 93.2(1), the set I0 is weakly admissible. Therefore, the set I := N \ I0 is
weakly admissible as well. Since the set I is non-empty and symmetric, hi × li ∈ I for
93. CRITERION OF ISOMORPHISM 387
some i. Take the smallest r with hr × lr ∈ I. By Proposition 93.3, there is an admissible

set J containing r. By Lemma 93.2(3), the union I0 ∪ J is also an admissible set. Let I1
be its symmetrization. Then the set I1 is admissible by Lemma 93.2(2). It is symmetric
and contains I0 properly because r ∈ I1 \ I0 .
So to finish, we must only prove Proposition 93.3.
Proof of Proposition 93.3. Multiplying the generic point morphism
Spec F (X) → X
by X × Y (on the left), we get a flat morphism
(X × Y )F (X) → X × Y × X .
This induces a surjective pull-back homomorphism
f : ChD (X̄ × Ȳ × X̄) → ChD (X̄ × Ȳ )
mapping each basis element of the form β1 × β2 × h0 to β1 × β2 and vanishing on the
remaining basis elements. Note that this homomorphism maps the subgroup of F -rational
cycles onto the subgroup of F (X)-rational cycles (cf. Corollary 57.10).
Since the quadrics ¡XF (X) and
¢ YF (X) are isotropic, the cycle cXY (N ) is F (X)-rational.
−1
Therefore, the set f cXY (N ) contains a rational cycle. Any cycle in this set has the
form
P
(93.4) c := cXY (N ) × h0 + α × β × γ ,
where the sum is taken over some homogeneous cycles α, β, γ with codim γ positive. In
the following, we assume that (93.4) is a rational cycle.
Let I and r be as in the statement of Proposition 93.3. Viewing the cycle (93.4) as a
correspondence from X̄ to Ȳ × X̄, we may take the composition c ◦ cXX (I). The result is
a rational cycle on X̄ × Ȳ × X̄ that (up to the addition of hd × hd ) is equal to
P
(93.5) c0 := cXY (I) × h0 + α × β × γ ,
where the sum is taken over some (other) homogeneous cycles α, β, γ with codim γ > 0
and codim α ≥ r. Take the pull-back of the cycle c0 with respect to the morphism
X̄ × Ȳ → X̄ × Ȳ × X̄, given by (x, y) 7→ (x, y, x), that is induced by the diagonal of X̄.
The result is a rational cycle on X̄ × Ȳ that is equal to
P
(93.6) c00 := cXY (I) + (α · γ) × β ,
where codim(α · γ) > r. It follows that c00 = cXY (J 0 ) for some set J 0 containing r.
Repeating the procedure with X and Y interchanged, we can find a set J 00 containing
r with the cycle cY X (J 00 ) rational. Then the set J := J 0 ∩ J 00 contains r and is admissible
as cXY (J) coincides (up to addition of hd × hd )) with the composition cXY (J 0 ) ◦ cY X (J 00 ) ◦
cXY (J 0 ) and a similar equality holds for cY X (J). ¤
The proof of Theorem 93.1 is now complete. ¤
Remark 93.7. By Theorem 27.3, an isomorphism of motives of odd-dimensional
quadrics gives rise to an isomorphism of the corresponding varieties. The question whether
for a given even n the conditions
n = dim ϕ = dim ψ and i0 (ϕL ) = i0 (ψL ) for any L
implies that ϕ and ψ are similar is answered positively if characteristic 6= 2 and n ≤ 6 in

[64], and negatively for all n ≥ 8 but 12 in [65]. It remains open for n = 12.
94. Indecomposable summands

In this section, Λ = Z/2Z and we work in the category CM(F, Z/2Z) of graded motives.
Let X be a smooth anisotropic projective quadric of dimension D. We write P for the
set of idempotents in ChD (X 2 ) = End M (X). We shall provide some information about
objects (X, p) (with p ∈ P ) in the category CM(F, Z/2Z). For such p (or, more generally,
for any element p ∈ ChD (X 2 )), let p̄ stand for the essence (as defined in §72) of the image
of p in the reduced Chow group Ch(X 2 ). We write [(X, p)] for the isomorphism class of
the motive (X, p).
Theorem 94.1. (1) Let p, p0 ∈ P . Then [(X, p)] = [(X, p0 )] if and only if p̄ = p̄0 .
Moreover, the image of the map
{[(X, p)]}p∈P → ChD (X 2 ) given by [(X, p)] 7→ p̄
is the group Che¯ D (X 2 ) of all D-dimensional essential cycles.
L
(2) Let p, p1 , p2 ∈ P . Then (X, p) ' (X, p1 ) (X, p2 ) if and only if p̄ is a disjoint
union of p̄1 and p̄2 , i.e., p̄1 and p̄2 don’t intersect and p̄ = p̄1 + p̄2 . In particular,
the motive (X, p) is indecomposable if and only if the cycle p̄ is minimal.
(3) For any p, p0 ∈ P , the motives (X, p) and (X, p0 ) are isomorphic to twists of each
other if and only if p̄ and p̄0 are derivatives of the same rational cycle. More
precisely, if i ≥ 0 then (X, p) ' (X, p0 )(i) if and only if p̄ = (h0 × hi ) · α and
p̄0 = (hi × h0 ) · α for some α ∈ ChD+i (X 2 ).
Proof. Let E/F be a field extension such that the quadric XE is split. The following
statements on projectors in End M (XE ) are easily checked: an element α ∈ ChD (XE2 ) is
a projector if and only if it is a linear combination of the elements hi × li and li0 × hi ,
i ∈ [0, d], where

 li for i < d
li0 = ld if D is divisible by 4
 d
h + ld otherwise
(cf. Exercise 68.3). Moreover, the condition (XE , α) ' (XE , α0 ) for two projectors α and
α0 means that α = α0 up to the terms with hd × ld and ld0 × hd , where these terms, if they
do not coincide, are equal to hd × ld for one of α and α0 and to ld0 × hd for the other.
(1): By Corollary 92.7, we have [(X, p)] = [(X, p0 )] if and only if [(X, p)E ] = [(X, p0 )E ].
Since the cycles pE and p0E are rational, it follows that [(X, p)E ] = [(X, p0 )E ] if and only if
pE = p0E . (Note by Exercise 92.8, we therefore get a canonical isomorphism of (X, p) and
(X, p0 ) once we have an isomorphism).
L Finally, pE = p0E if and only if p̄ = p̄L
0
.
(2): We have (X, p) ' (X, p1 ) (X, p2 ) if and only if (X, p)E ' (X, p1 )E (X, p2 )E if
and only if pE is a disjoint union of (p1 )E and (p2 )E if and only if p̄ is a disjoint union of
p̄1 and p̄2 .
(3): A correspondence α ∈ ChD+i (X 2 ) determines an isomorphism (X, p)E → (X, p0 )(i)E
if and only if p̄ = (h0 × hi ) · ᾱ and p̄0 = (hi × h0 ) · ᾱ. ¤
94. INDECOMPOSABLE SUMMANDS 389
Corollary 94.2. The motive of any anisotropic smooth projective quadric X decom-
poses into a direct sum of indecomposable summands. Moreover, such a decomposition is
unique and the number of summands coincides with the number of the minimal cycles in
ChD (X 2 ), where D = dim X.
Exercise 94.3 (Rost motives). Let π be an anisotropic n-fold Pfister form. Show the
following:
(1) The decomposition of the motive of the projective quadric of π into a sum of indecom-
L n−1
posable summands has the form 2i=0 −1 Rπ (i) for some motive Rπ uniquely determined
by π. The motive Rπ is called the Rost motive associated to π.
(2) For any splitting field extension E/F of π, we have
(Rπ )E ' Z/2Z ⊕ Z/2Z(2n−1 − 1).
(3) The motive of the quadric given by any 1-codimensional subform of π decomposes as
L2n−1 −2
i=0 Rπ (i).
(4) Let ϕ be a (2n−1 + 1)-dimensional non-degenerate subform of π. Find a smooth
projective quadric X such that the motive of the quadric of ϕ decomposes as M (X)(1) ⊕
Rπ . Finally, reprove all this for motives with integral coefficients.
Theorems 93.1 and 94.1 hold in the more general case of motives with integral coeffi-
cients. This was done by Vishik in [140].
Appendices
95. Formally real fields
In this section, we review the Artin-Schreier theory of formally real fields. These
results and their proofs, may be found in the books by Lam [90] and Scharlau [125].
Let F be a field, P ⊂ F a subset. We say that P is a preordering of F if P satisfies
all of the following:
X
P + P ⊂ P, P · P ⊂ P, −1 ∈/ P, and F 2 ⊂ P.
A preordering P of F is called an ordering if in addition
F = P ∪ −P.
A field F is called formally real if
e
D(∞h1i) := {x ∈ F | x is a sum of squares in F }
is a preordering of F , equivalently if −1 is not a sum of squares in F , i.e., the polynomial
t21 + · · · + t2n has no nontrivial zero over F for any (positive) integer n. Clearly, if F is
formally real then the characteristic of F must be zero. (If char F 6= 2 then F is not
e
formally real if and only if F = D(∞h1i).) One checks that a preordering is an ordering if
and only if it is maximal with respect to set inclusion in the set of preorderings of F . By
Zorn’s lemma, maximal preorderings and therefore orderings exist for F if it is formally
real. In particular, a field F is formally real if and only if the space of orderings on F ,
X(F ) := {P | P is an ordering of F }
e
is not empty. Every P ∈ X(F ) (if any) contains the preordering D(∞h1i). Let P ∈ X(F )
and 0 6= x ∈ F . If x ∈ P then x (respectively, −x) is called positive (respectively, negative)
with respect to P and we write x >P 0 (respectively, x <P 0). Elements that are positive
(respectively negative) with respect to all orderings of F (if any) are called totally positive
(respectively, totally negative). In fact, we have
e T
Proposition 95.1. Suppose that F is formally real. Then D(∞h1i) = P ∈X(F ) P ,
i.e., a nonzero element of F is totally positive if and only if it is a sum of squares.
e
It follows that a formally real field has precisely one ordering if and only if D(∞h1i)
is an ordering in F , e.g., Q or R. The field of real numbers even has R as an ordering.
2
A formally real field F having F 2 as an ordering is called euclidean. For such a field
every element is either a square or the negative of a square. For example, the field of real
constructible numbers is euclidean.
A formally real field is called real closed if it has no proper algebraic extension that
is formally real. If F is such a field then it must be euclidean. Let K/F be an algebraic
field extension with K real closed. Then K 2 ∩ F is an ordering on F .
Let Q ∈ X(K). The pair (K, Q) is called an ordered field. Let K/F be a field extension
with K formally real. If P ∈ X(F ) satisfies P = Q ∩ F then (K, Q)/(F, P ) is called an
extension of ordered fields and Q is called an extension of P . If, in addition, K/F is
algebraic and there exist no extension (L, R)/(K, Q) with L/K non-trivial algebraic, we
call (K, Q) a real closure of (F, P ).
96. THE SPACE OF ORDERINGS 393
Proposition 95.2. (Cf. [125, Th. 3.1.14].) If (K, Q) is a real closure of (F, P ) then
K is real closed and Q = K 2 .
The key to proving this is
Theorem 95.3. (Cf. [125, Th. 3.1.9].) Let (F, P ) be an ordered field.
√
(1) Let d ∈ F and K = F ( d). Then there exists an extension of P to K if and
only if d ∈ P .
(2) If K/F is finite of odd degree then there exists an extension of P to K.
The main theorem of Artin-Schreier Theory is
Theorem 95.4. (Cf. [90, Th. VIII.2.8] or [125, Th. 3.1.13 and Th. 3.2.8].) Every
2
ordered field (F, P ) has a real closure (F , F ) and this real closure is unique up to a
canonical F -isomorphism and this isomorphism is order-preserving.
Because of the last results, if we fix an algebraic closure Fe of a formally real field F
2
and P ∈ X(F ) then there exists a unique real closure (F , F ) of (F, P ) with F ⊂ Fe. We
denote F by FP .
96. The space of orderings
We view the space of orderings X(F ) on a field F as a subset of the space of functions
×
{±1}F by the embedding
×
X(F ) → {±1}F via P 7→ (signP : x 7→ signP x)
×
(the sign of x in F rel P ). Giving {±1} the discrete topology, we have {±1}F is Hausdorff
and by Tychonoff’s Theorem compact. The collection of clopen (i.e., open and closed)
sets given by
×
(96.1) Hε (a) := {g ∈ {±1}F | g(a) = −ε}
× ×
for a ∈ F × and ε ∈ {±1} forms a subbase for the topology of {±1}F , hence {±1}F
×
is also totally disconnected. Consequently, {±1}F is a boolean space (i.e., a compact
totally disconnected Hausdorff space). Let X(F ) have the induced topology arising from
×
the embedding f : X(F ) → {±1}F .
Theorem 96.2. X(F ) is a boolean space.
× ×
Proof. It suffices to show that X(F ) is closed in {±1}F . Let s ∈ {±1}F \f (X(F )).
First suppose that s is the constant function ε. Then the clopen set Hε (ε) is disjoint from
f (X(F )) and contains s, so separates s from f (X(F )). So assume that s is not a constant
function hence is surjective. Since s−1 (1) is not an ordering on F , there exist a, b ∈ F ×
such that s(a) = 1 = s(b) (i.e., a, b are “positive”) but either s(a + b) = −1 or s(ab) = −1.
Let c = ab if s(ab) = −1 otherwise let c = a + b. As there cannot be an ordering in which
a and b are positive but c negative, H1 (−a) ∩ H1 (−b) ∩ H−1 (−c) is disjoint from f (X(F ))
and contains s, so separates s from f (X(F )). ¤
394
×
As we are identifying X(F ) with its image in {±1}F , we see that the collection of
sets
H(a) = HF (a) := H1 (a) ⊂ X(F ), a ∈ F × ,
forms a subbasis of clopen sets for the topology of X(F ) called the Harrison subbasis. So
H(a) is the set of orderings on which a is negative. It follows that the collection of sets
n
\
H(a1 , . . . , an ) = HF (a1 , . . . , an ) := H(ai ), a1 , . . . , an ∈ F ×
i=1
forms a basis for the topology of X(F ).
97. Cn -fields
We call a homogeneous polynomial of (total) degree d a d-form. A field F is called a

Cn -field if every d-form over F in at least dn + 1 variables has a non-trivial zero over F .
For example, a field is algebraically closed if and only if it is a C0 -field. Every finite
field is a C1 -field by the Chevalley-Warning Theorem (cf. [127], I.2, Theorem 3).
An n-form in n-variables over F is called a normic form if it has no non-trivial zero.
For example, let E/F be a finite field extension of degree n. Let {x1 , . . . , xn } be an
F -basis for E. Then the norm form of the extension E/F is the polynomial
NE(t1 ,...,tn )/F (t1 ,...,tn ) (t1 x1 + · · · + tn xn )
over F in the variables t1 , . . . , tn is of degree n and has no nontrivial zero, hence is normic
(the reason for the name).
Lemma 97.1. Let F be a non algebraically closed field. Then there exist normic forms
of arbitrarily large degree.
Proof. There exists a normic form ϕ of degree n for some n > 1. Having defined a
normic form ϕs of degree ns , let
ϕs+1 := ϕ(ϕs |ϕs | . . . |ϕs ).
This notation means that new variables are to be used after each occurrence of |. The
form ϕs+1 of degree ns+1 has no non-trivial zero. ¤
Theorem 97.2. Let F be a Cn -field and let f1 , . . . , fr be d-forms in N common vari-
ables. If N > rdn then the forms have a common non-trivial zero in F .
Proof. Suppose first that n = 0 (i.e., F is algebraically closed) or d = 1. As N > r,
it follows from [130, Ch. I, §6.2, Prop.] that the forms have a common non-trivial zero
over F .
So we may assume that n > 0 and d > 1. By Lemma 97.1, there exists a normic form
ϕ of degree at least r. We define a sequence of forms ϕi , i ≥ 1, of degree di in Ni variables
as follows. Let ϕ1 = ϕ. Assuming that ϕi is defined let
ϕi+1 = ϕ(f1 , . . . , fr | f1 , . . . , fr | . . . | f1 , . . . , fr | 0, . . . , 0),
where zeros occur in < r places. The forms fi between two consecutive signs | have the
same sets of variables.
97. Cn -FIELDS 395
If x ∈ R let [x] denote the largest integer ≤ x. We have

hN i
i
(97.3) di+1 = ddi and Ni+1 = N .
r
Note that since N > rdn ≥ 2r we have Ni → ∞ as i → ∞.
Set
r h Ni i
(97.4) αi = .
Ni r
We have αi → 1 as i → ∞. It follows from (97.3) and (97.4) that
Ni+1 Ni αi N
n
= n · n.
di+1 di rd
αi N
Since N > rdn and αi → 1, there is β > 1 and an integer s such that > β if i ≥ s.
rdn
Therefore, we have
Ni+1 Ni
n
> n ·β
di+1 di
n
if i ≥ s. It follows that Nk > dk for some k. As F is a Cn -field, the form ϕk has a nontrivial
zero. Choose the smallest k with this property. By definition of ϕk , a nontrivial zero of
ϕk gives rise to a nontrivial common zero of the forms f1 , . . . , fr . ¤
Corollary 97.5. Let F be a Cn -field and K/F an algebraic field extension. Then K
is a Cn -field.
Proof. Let f be a d-form over K in N variables with N > dn . The coefficients of f
belong to a finite field extension of F , so we may assume that K/F is a finite extension.
Let {x1 , . . . , xr } be an F -basis for K. Choose variables tij , i = 1, . . . , N , j = 1, . . . r over
F and set
ti = ti1 x1 + · · · + tir xr
for every i. Then
f (t1 , . . . , tN ) = f1 (tij )x1 + . . . fr (tij )xr
for some d-forms fj in rN variables. Since rN > rdn , it follows from Theorem 97.2 that
the forms fj have a nontrivial common zero over F which produces a nontrivial zero of f
over K. ¤
Corollary 97.6. Let F be a Cn -field. Then F (t) is a Cn+1 -field.
Proof. Let f be a d-form in N variables over F (t) with N > dn+1 . Clearing denom-
inators of the coefficients of f we may assume that all the coefficients are polynomials in
t. Choose variables tij , i = 1, . . . , N , j = 0, . . . , m for some m and set
ti = ti0 + ti1 t + · · · + tim tm
for every i. Then
f (t1 , . . . , tN ) = f0 (tij )t0 + · · · + fdm+r (tij )tdm+r
for some d-forms fj in N (m + 1) variables over F and r = degt (f ). Since N > dn+1 , one
can choose m such that N (m + 1) > (dm + r + 1)dn . By Theorem 97.2, the forms fj have
a nontrivial common zero over F which produces a nontrivial zero of f over F (t). ¤
396
Corollaries 97.5 and 97.6 yield

Theorem 97.7. Let F be a Cn -field and K/F a field extension of transcendence degree
m. Then K is a Cn+m -field.
As algebraically closed field are C0 -fields, the theorem shows that a field of transcen-
dence degree n over an algebraically closed field is a Cn -field. In particular, we have the
classical Tsen Theorem:
Theorem 97.8. If F is algebraically closed and K/F is a field extension of transcen-
dence degree 1 then the Brauer group Br(K) is trivial.
Proof. Let A be a central division algebra over K of degree d > 1. The reduced
norm form Nrd of D is a form of degree d in d2 variables. By Theorem 97.7, K is a
C1 -field, hence Nrd has a nontrivial zero, a contradiction. ¤
98. Algebras
For more details see [86] and [53].
98.A. Semisimple, separable and étale algebras. Let F be a field. A finite

dimensional (associative, unital) F -algebra A is called simple if A has no nontrivial (two-
sided) ideals. By Wedderburn’s theorem, every simple F -algebra is isomorphic to the
matrix algebra Mn (D) for some n and a division F -algebra D uniquely determined by A
up to isomorphism over F .
An F -algebra A is called semisimple if A is isomorphic to a (finite) product of simple
algebras.
An F -algebra A is called separable if the L-algebra AL := A ⊗F L is semisimple for
every field extension L/F . This is equivalent to A being a finite product of the matrix
algebras Mn (D), where D is a division F -algebra with center a finite separable field
extension of F . Separable algebras satisfy the following descent condition:
Fact 98.1. If A is an F -algebra and E/F is a field extension then A is separable if
and only if AE is separable as an E-algebra.
Let A be a finite dimensional commutative F -algebra. If A is separable, it is called
étale. Consequently, A is étale if and only if A is a finite product of finite separable field
extensions of F . An étale F -algebra A is called split if A is isomorphic to a product of
several copies of F . Let A be a commutative (associative, unital) F -algebra. The determi-
nant (respectively, the trace) of the linear endomorphism of A given by left multiplication
by an element a ∈ A is called the norm NA (a) (respectively, the trace TrA (a)). We have
TrA (a + a0 ) = TrA (a) + TrA (a0 ) and NA (aa0 ) = NA (a) NA (a0 ) for all a, a0 ∈ A. Every a ∈ A
satisfies the characteristic polynomial equation
an − TrA (a)an−1 + · · · + (−1)n NA (a) = 0
where n = dim A.
98. ALGEBRAS 397
98.B. Quadratic algebras. A quadratic algebra A over F is an F -algebra of dimen-

sion 2. A quadratic algebra is necessarily commutative. Every element a ∈ A satisfies the
quadratic equation
(98.2) a2 − TrA (a)a + NA (a) = 0.
For every a ∈ A, set ā := TrA (a) − a. We have aa0 = āā0 for all a, a0 ∈ A. Indeed, since
dim A = 2, it suffices to check the equality when a ∈ F and a0 ∈ F (this is obvious)
and a0 = a (it follows from the quadratic equation). Thus the map a 7→ ā is an algebra
automorphism of A of exponent 2. We have
TrA (a) = a + ā and NA (a) = aā.
A quadratic F -algebra A is étale if A is either a quadratic separable field extension of
F or A is split, i.e., is isomorphic to F × F .
Let A and B be two quadratic étale F -algebras. The subalgebra A ? B of the tensor
product A ⊗F B consisting of all elements stable under the automorphism of A ⊗F B
defined by x ⊗ y 7→ x̄ ⊗ ȳ is also a quadratic étale F -algebra. The operation ? on
quadratic étale F -algebras yields a (multiplicative) group structure on the set Ét2 (F ) of
isomorphisms classes [A] of quadratic étale F -algebras A. Thus [A] · [B] = [A ? B]. Note
that Ét2 (F ) is an abelian group of exponent 2.
Example 98.3. If char F 6= 2, every quadratic étale F -algebra is isomorphic to
Fa := F [t]/(t2 − a)
for some a ∈ F × . Let j be the class of t in Fa . For every u = x + yj, we have
ū = x − yj, Tr(u) = 2x, andN(u) = x2 − ay 2 .
The assignment a 7→ [Fa ] give rise to an isomorphism F × /F ×2 ∼
= Ét2 (F ).
Example 98.4. If char F = 2, every quadratic étale F -algebra is isomorphic to
Fa := F [t]/(t2 + t + a)
for some a ∈ F . Let j be the class of t in Fa . For every u = x + yj, we have
ū = x + y + yj, Tr(u) = y, N(u) = x2 + xy + ay 2 .
and
The assignment a 7→ [Fa ] induces an isomorphism F/ Im ℘ ∼
= Ét2 (F ), where the Artin-
Schreier map ℘ : F → F is defined by ℘(x) = x2 + x.
98.C. Brauer group. An F -algebra A is called central if F 1 coincides with the

center of A. A central simple F -algebra A is called split if A ∼ = Mn (F ) for some n.
Two central simple F -algebras A and B are called Brauer equivalent if Mn (A) ∼ =
Mm (B) for some n and m. For example, all split F -algebras are Brauer equivalent.
The set Br(F ) of all Brauer equivalence classes of central simple F -algebras is a torsion
abelian group with respect to the tensor product operation A ⊗F B, called the Brauer
group of F . The identity element of Br(F ) is the class of split F -algebras.
The class of a central simple F -algebra A will be denoted by [A] and the product of [A]
and [B] in the Brauer group, represented by the tensor product A ⊗F B, will be denoted
by [A] · [B].
398
The inverse class of A in Br(F ) is given by the class of the opposite algebra Aop . The
order of [A] in Br(F ) is called the exponent of A and will be denoted by exp(A). In
particular, exp(A) divides 2 if and only if Aop ∼
= A, i.e., A has an anti-automorphism.
For an integer m, we write Brm (F ) for the subgroup of all classes [A] ∈ Br(F ) such
that [A]m = 1.
Let A be a central simple algebra over F and L/F a field extension. Then AL is a
central simple algebra over L. (In particular, every central simple F -algebra is separable.)
The correspondence [A] 7→ [AL ] gives rise to a group homomorphism
rL/F : Br(F ) → Br(L).
We set Br(L/F ) := ker rL/F . The class A is said to be split over L (and L/F is called a
splitting field extension of A) if the algebra AL is split, equivalently [A] ∈ Br(L/F ).
A central simple F -algebra A is isomorphic√to Mk (D) for√a central division F -algebra
D, unique up to isomorphism. The integers dim D and dim A are called the index
and the degree of A respectively and denoted by ind(A) and deg(A).
Fact 98.5. Let A be a central simple algebra over F and L/F a finite field extension.
Then
ind(AL ) | ind(A) | ind(AL ) · [L : F ].
Corollary 98.6. Let A be a central simple algebra over F and L/F a finite field
extension. Then
(1) If L is a splitting field of A then ind(A) divides [L : F ].
(2) If [L : F ] is relatively prime to ind(A) then ind(AL ) = ind(A).
Fact 98.7. Let A be a central division algebra over F .
(1) A subfield K ⊂ A is maximal if and only if [K : F ] = ind(A). In this case K is
a splitting field of A.
(2) Every splitting field of A of degree ind(A) over F can be embedded into A over
F as a maximal subfield.
98.D. Severi-Brauer varieties. Let A be a central simple F -algebra of degree n.
Let r be an integer dividing n. The generalized Severi-Brauer variety SBr (A) of A is the
variety of right ideals of dimension rn in A [86, 1.16]. We simply write SB(A) for SB1 (A).
If A is split, i.e., A = End(V ) for a vector space V of dimension n, every right
ideal I in A of dimension rn has the form I = Hom(V, U ) for a uniquely determined
subspace U ⊂ V of dimension r. Thus the correspondence I 7→ U yields an isomorphism
SBr (A) ∼= Grr (V ), where Grr (V ) is the Grassmannian variety of r-dimensional subspaces
in V . In particular, SB(A) ∼ = P(V ).
Proposition 98.8. [86, Prop. 1.17] Let A be a central simple F -algebra, r an integer
dividing deg(A). Then the Severi-Brauer variety X = SBr (A) has a rational point over
an extension L/F if and only if ind(AL ) divides r. In particular, SB(A) has a rational
point over L if and only if A is split over L.
Let V1 and V2 be vector spaces over F of finite dimension. The Segre closed embedding
is the morphism
P(V1 ) × P(V2 ) → P(V1 ⊗F V2 )
98. ALGEBRAS 399
taking a pair of lines U1 and U2 in V1 and V2 respectively to the line U1 ⊗F U2 in V1 ⊗F V2 .

Example 98.9. The Segre embedding identifies P1F × P1F with a projective quadric in
P3F .
The Segre embedding can be generalized as follows. Let A1 and A2 be two central
simple algebras over F . Then the correspondence (I1 , I2 ) 7→ I1 ⊗ I2 yields a closed
embedding
SB(A1 ) × SB(A2 ) → SB(A1 ⊗F A2 ).
98.E. Quaternion algebras. Let L/F be a Galois quadratic field extension with
Galois group {e, g} and b ∈ F × . The F -algebra Q = (L/F, b) := L ⊕ Lj, where the
symbol j satisfies j 2 = b and jl = g(l)j for all l ∈ L, is central simple of dimension
4 and is called a quaternion algebra. We have Q is either split, i.e., isomorphic to the
matrix algebra M2 (F ) or a division algebra. The algebra Q carries a canonical involution
: Q → Q satisfying j̄ = −j and ¯l = g(l) for all l ∈ L.
Using the canonical involution, we define the linear reduced trace map
Trd : Q → F by Trd(q) = q + q̄,
and quadratic reduced norm map
Nrd : Q → F by Nrd(q) = q · q̄.
An element q ∈ Q is called a pure quaternion if Trd(q) = 0, or equivalently, q̄ = −q.
Denote by Q0 the 3-dimensional subspace of all pure quaternions. We have Nrd(q) = −q 2
for any q ∈ Q0 .
Proposition 98.10. Every central division algebra of dimension 4 is isomorphic to a
quaternion algebra.
Proof. Let L ⊂ Q be a separable quadratic subfield. By the Skolem-Noether Theo-
rem, the only nontrivial automorphism g of L over F extends to an inner automorphism
of Q, i.e., there is j ∈ Q× such that jlj −1 = g(l) for all l ∈ L. Clearly, Q = L ⊕ Lj
and j 2 commutes with j and L. Hence j 2 belongs to the center of Q, i.e., b := j 2 ∈ F × .
Therefore, Q is isomorphic to the quaternion algebra Q = (L/F, b). ¤
Example 98.11. If char F 6= 2, a separable quadratic subfield L of a quaternion
algebra Q = (L/F, b) is of the form L = F (i) with i2 = a ∈ F × . Hence Q has a basis
{1, i, j, k := ij} with multiplication table
i2 = a, j 2 = b, ji + ij = 0,
some¶b ∈ F × . We shall denote the algebra generated by i and j with these relations
for µ
a, b
by .
F
The space of pure quaternions has {i, j, k} as a basis. For every q = x + yi + zj + wk
with x, y, z, w ∈ F , we have
q̄ = x − yi − zj − wk, Trd(q) = 2x, and Nrd(q) = x2 − ay 2 − bz 2 + abw2 .
400
Example 98.12. If char F = 2, a separable quadratic subfield L of a quaternion

algebra Q = (L/F, b) is of the form L = F (s) with s2 + s + c = 0 for some c ∈ F .
Set i = sj/b. We have i2 = a := c/b. Hence Q has a basis {1, i, j, k := ij} with the
multiplication table
(98.13) i2 = a, j 2 = b, ji + ij = 1.
· ¸
a, b
We shall denote by the algebra given by the generators i and j and the relations
F
(98.13). Note that this is also a quaternion algebra (in fact split) when b = 0.
The space of pure quaternions has {1, i, j} as a basis. For every q = x + yi + zj + wk
with x, y, z, w ∈ F , we have
q̄ = (x+w)+yi+zj +wk, Trd(q) = w, and Nrd(q) = x2 +ay 2 +bz 2 +abw2 +xw+yz.
The classes of quaternion F -algebras satisfy the following relations in Br(F ):

Fact 98.14. Suppose that char F 6= 2. Then
µ 0 ¶ µ ¶ µ 0 ¶ µ ¶ µ ¶ µ 0¶
aa , b a, b a ,b a, bb0 a, b a, b
(1) = · and = · .
F F F F F F
µ ¶ µ ¶
a, b b, a
(2) = .
F F
µ ¶2
a, b
(3) = 1.
F
µ ¶
a, b
(4) = 1 if and only if a is a norm of the quadratic étale extension Fb /F .
F
Fact 98.15. Suppose that char F = 2. Then

· ¸ · ¸ · 0 ¸ · ¸ · ¸ · 0¸
a + a0 , b a, b a ,b a, b + b0 a, b a, b
(1) = · and = · .
F F F F F F
· ¸ · ¸ · ¸
ab, c bc, a ca, b
(2) · · = 1.
F F F
· ¸ · ¸
a, b b, a
(3) = .
F F
· ¸2
a, b
(4) = 1.
F
· ¸
a, b
(5) = 1 if and only if a is a norm of the quadratic étale extension Fab /F .
F
We shall need the following properties of quaternion algebras.

98. ALGEBRAS 401
µ ¶ µ ¶
a, b c, d
Lemma 98.16. (Chain Lemma) Let and be isomorphic quaternion
F F
algebras
µ ¶ over a field F of characteristic
µ ¶ not 2. Then there is an e ∈ F × satisfying
a, b ³ a, e ´ ³ c, e ´ c, d
' ' ' .
F F F F
Proof. Note that if x and y are pure quaternions in a quaternion algebra Q that are
orthogonal
µ 2 2 ¶ with respect µ to¶the reduced trace bilinear form, i.e., Trd(xy) = 0, then Q '
x ,y a, b
. Let Q = . By assumption, there are pure quaternions x, y satisfying
F F
x2 = a and y 2³ = c.´ Choose 2
³ c, e á pure quaternion z orthogonal to x and y. Setting e = z ,
a, e
we have Q ' ' . ¤
F F
Lemma 98.17. Let Q be a quaternion algebra over a field F of characteristic 2. Suppose
2
that Q
· is ¸split by a purely inseparable field extension K/F such that K ⊂ F . Then
a, b
Q∼ = with a ∈ K 2 .
F
√
Proof. First suppose that K = F ( a) is a quadratic extension of F . By Fact 98.7,
we know that K can be embedded into Q. Therefore there exists an i ∈ Q \ F such that
i2 = a ∈ K 2 . Note that i is a pure quaternion in Q0 \ F . The linear map Q0 → F taking
· ¸
0 ∼ a, b
an x to ix + xi is nonzero, hence there is a j ∈ Q such that ij + ji = 1. Hence, Q =
F
where b = j 2 . · ¸
c, d
In the general case, write Q = . By Property (5), we have c = x2 + xy + cdy 2
F
for some x, y ∈ K. Since x2 and y 2 belong to F , we have xy ∈ F . Hence the extension
E = F (x, y) splits Q and [E : F ] ≤ 2. The statement now follows from the first part of
the proof. ¤
Let σ be an automorphism of a ring R. Denote by R[t, σ] the ring of σ-twisted poly-

nomials in the variable t with multiplication defined by tr = σ(r)t for all r ∈ R. For
example, if σ is the identity then R[t, σ] is the ordinary polynomial ring R[t] over R.
Observe that if R has no zero divisors then neither does R[t, σ].
Example 98.18. Let A be a central division algebra over a field F . Consider an

automorphism σ of the polynomial ring A[x] defined by σ(a) = a for all a ∈ A and
½
−x if char F 6= 2
σ(x) =
x + 1 if char F = 2.
Let B be the quotient ring of A[x][t, σ]. The ring B is a division algebra over its center
E where
½
F (x2 , t2 ) if char F 6= 2
E= 2 2
F (x + x, t ) if char F = 2.
402
Moreover, B = A ⊗F Q, where Q is a quaternion algebra over E satisfying

 µ 2 2¶

 x ,t

 if char F 6= 2
 E
Q= · 2 ¸

 (x + x)/t2 2
, t


 if char F = 2.
E
Iterating the construction in Example 98.18 yields the following
Proposition 98.19. For any field F and integer n ≥ 1, there is a field extension L/F
and a central division L-algebra that is a tensor product of n quaternion algebras.
We now study interactions between two quaternion algebras.
Theorem 98.20. Let Q1 and Q2 be division quaternion algebras over F . Then the
following conditions are equivalent:
(1) The tensor product Q1 ⊗F Q2 is not a division algebra.
(2) Q1 and Q2 have isomorphic separable quadratic subfields.
(3) Q1 and Q2 have isomorphic quadratic subfields.
Proof. (1) ⇒ (2): Write X1 , X2 and X for Severi-Brauer varieties of Q1 , Q2 and
A := Q1 ⊗F Q2 respectively. The morphism X1 × X2 → X taking a pair of ideals I1 and
I2 to the ideal I1 ⊗ I2 identifies X1 × X2 with a twisted form of a 2-dimensional quadric
in X (cf. §98.D).
Let Y be the generalized Severi-Brauer variety of rank 8 ideals in A. A rational point
of Y , i.e., a right ideal J ⊂ A of dimension 8, defines the closed curve CJ in X comprising
of all ideals of rank 4 contained in J. In the split case, Y is the Grassmannian variety of
planes and CJ is the projective line (the projective space of the plane corresponding to
J) intersecting properly the quadric X1 × X2 in two points. Thus there is a nonempty
open subset U ⊂ Y with the following property: for any rational point J ∈ U , we have
CJ ∩ (X1 × X2 ) = {x}, where x is a point of degree 2 with residue field L a separable
quadratic field extension of F . By assumption, there is a right ideal I ⊂ A of dimension
8, i.e., Y (F ) 6= ∅. The algebraic group G of invertible elements of A acts transitively on
Y , i.e., the morphism G → Y taking an a to the ideal aI is surjective. As rational point
of G are dense in G, we have rational points of Y are dense in Y . Hence U possesses a
rational point J.
As X1 (L) × X2 (L) = (X1 × X2 )(L) 6= ∅, it follows that the field L split both Q1 and
Q2 and therefore L is isomorphic to quadratic subfields in Q1 and Q2 .
(3) ⇒ (1): Let L/F be a common quadratic subfield of both Q1 and Q2 . It follows that
Q1 and Q2 and hence A are split by L. It follows from Corollary 98.6 that ind(A) ≤ 2,
i.e., A is not a division algebra. ¤
99. Galois cohomology

For more details see [129].
99. GALOIS COHOMOLOGY 403
99.A. Galois modules and Galois cohomology groups. Let Γ be a profinite

group and M a (left) discrete Γ-module. For any n ∈ Z, let H n (Γ, M ) denote the nth
cohomology group of Γ with coefficients in M . In particular, H n (Γ, M ) = 0 if n < 0 and
© ª
H 0 (Γ, M ) = M Γ := m ∈ M | γm = m for all γ ∈ Γ ,
the subgroup of Γ-invariant elements of M .
An exact sequence 0 → M 0 → M → M 00 → 0 gives rise to an infinite long exact
sequence of cohomology groups
0 → H 0 (Γ, M 0 ) → H 0 (Γ, M ) → H 0 (Γ, M 00 ) → H 1 (Γ, M 0 ) → H 1 (Γ, M ) → . . .
Let F be a field. Denote by ΓF the absolute Galois group of F , i.e., the Galois group
of a separable closure Fsep of the field F over F . A discrete ΓF -module is called a Galois
module over F . For a Galois module M over F , we write H n (F, M ) for the cohomology
group H n (ΓF , M ).
Example 99.1. (1) Every abelian group A can be viewed as a Galois module over
F with trivial action. We have H 0 (F, A) = A and H 1 (F, A) = Homc (ΓF , A), the group
of continuous homomorphisms (where A is viewed with discrete topology). In particular,
H 1 (F, A) is trivial if A is torsion-free, e.g., H 1 (F, Z) = 0.
The group H 1 (F, Q/Z) = Homc (ΓF , Q/Z) is called the character group of ΓF and will
be denoted by χ(ΓF ).
The cohomology group H n (F, M ) is torsion for every Galois module M and any n ≥ 1.
Since the group Q is uniquely divisible, we have H n (F, Q) = 0 for all n ≥ 1. The
cohomology exact sequence for the short exact sequence of Galois modules with trivial
action
0 → Z → Q → Q/Z → 0
then gives an isomorphism H n (F, Q/Z) → H n+1 (F, Z) for any n ≥ 1. In particular,
∼
H 2 (F, Z) ∼
= χ(ΓF ).
Let m be a natural integer. The cohomology exact sequence for the short exact
sequence
0→Z− → Z → Z/mZ → 0
m
gives an isomorphism of H 1 (F, Z/mZ) with the subgroup χm (ΓF ) of characters of exponent
m.
(2) The cohomology groups H n (F, Fsep ) with coefficients in the additive group Fsep
are trivial if n > 0. If char F = p > 0, the cohomology exact sequence for the short exact
sequence
℘
0 → Z/pZ → Fsep →− Fsep → 0,
where ℘ is the Artin-Schreier map defined by ℘(x) = xp −x, yields canonical isomorphisms

 Z/pZ if n = 0
H (F, Z/pZ) =
n ∼ F/℘(F ) if n = 1

0 otherwise.
In fact, H n (F, M ) = 0 for all n ≥ 2 and every Galois module M over F of characteristic
p satisfying pM = 0.
404
(3) We have the following canonical isomorphisms for the cohomology groups with
×
coefficients in the multiplicative group Fsep :

 F× if n = 0
H (F, Fsep ) ∼
n ×
= 1 if n = 1 (Hilbert Theorem 90)
 Br(F ) if n = 2.
(4) The group µm = µm (Fsep ) of mth roots of unity in Fsep is a Galois submodule of
×
Fsep . We have the following exact sequence of Galois modules:
× × × ×m
(99.2) 1 → µm → Fsep → Fsep → Fsep /Fsep → 1,
where the middle homomorphism takes x to xm .
× ×m
If m is not divisible by char F , we have Fsep /Fsep = 1. Therefore, the cohomology
exact sequence (99.2) yields isomorphisms

 µm (F ), if n = 0
n ∼
H (F, µm ) = × ×m
F /F , if n = 1

Brm (F ), if n = 2.
We shall write (a)m or simply (a) for the element of H 1 (F, µm ) corresponding to a coset
aF ×m in F × /F ×m .
If p = char F > 0, we have µp (Fsep ) = 1 and the cohomology exact sequence (99.2)
gives an isomorphism
¡ ×p ∼
¢
H 1 F, Fsep
×
/Fsep = Brp (F ).
Example 99.3. Let ξ ∈ χ2 (ΓF ) be a nontrivial character. Then ker(ξ) is a subgroup
of ΓF of index 2. By Galois theory, it corresponds to a Galois quadratic field extension
∼
Fξ /F . The correspondence ξ 7→ Fξ gives rise to an isomorphism χ2 (ΓF ) → Ét2 (F ).
99.B. Cup-products. Let M and N be Galois modules over F . Then the tensor
product M ⊗Z N is also a Galois modules via the diagonal ΓF -action. There is a pairing
H m (F, M ) ⊗ H n (F, N ) → H m+n (F, M ⊗Z N ), α ⊗ β 7→ α ∪ β
called the cup-product. When m = n = 0 the cup-product coincides with the natural
homomorphism M ΓF ⊗ N ΓF → (M ⊗Z N )ΓF .
Fact 99.4. [26, Ch. IV, §7] Let 0 → M 0 → M → M 00 → 0 be an exact sequence of
Galois modules over F . Suppose that for a Galois module N the sequence
0 → M 0 ⊗Z N → M ⊗Z N → M 00 ⊗Z N → 0
is exact. Then the diagram
∪
H n (F, M 00 ) ⊗ H m (F, N ) −−−→ H n+m (F, M 00 ⊗Z N )
 
∂⊗idy
 
y∂
∪
H n+1 (F, M 0 ) ⊗ H m (F, N ) −−−→ H n+m+1 (F, M 0 ⊗Z N )
is commutative.
99. GALOIS COHOMOLOGY 405
Example 99.5. The cup-product

H 0 (F, Fsep
×
) ⊗ H 2 (F, Z) → H 2 (F, Fsep
×
)
yields a pairing
F × ⊗ Ét(F ) → Br2 (F ).
µ ¶
a, b
If char F 6= 2, we have a ∪ [Fb ] = for all a, b ∈ F × . In the case that char F = 2,
· ¸ F
a, b
we have a ∪ [Fab ] = for all a ∈ F × and b ∈ F .
F
Suppose that char F 6= 2. Then µ2 ' Z/2Z. The cup-product
H 1 (F, Z/2Z) ⊗ H 1 (F, Z/2Z) → H 2 (F, Z/2Z)
gives rise to a pairing
2 2
F × /F × ⊗ F × /F × → Br2 (F ).
µ ¶
a, b
We have (a) ∪ (b) = for all a, b ∈ F × . In particular, (a) ∪ (1 − a) = 0 for every
F
a 6= 0, 1 by Fact 98.14(4).
99.C. Restriction and corestriction homomorphisms. Let M be a Galois mod-
ule over F and K/F an arbitrary field extension. Separable closures of F and K can be
chosen so that Fsep ⊂ Ksep . The restriction then yields a continuous group homomor-
phism ΓK → ΓF . In particular, M has the structure of a discrete ΓK -module and we have
the restriction homomorphism
rK/F : H n (F, M ) → H n (K, M ).
If K/F is a finite separable field extension then ΓK is an open subgroup of finite index
in ΓF . For every n ≥ 0 there is a natural corestriction homomorphism
cK/F : H n (K, M ) → H n (F, M ).
P
In the case n = 0, the map cK/F : M ΓK → M ΓF is given by x → γ(x) where the sum
is over a left transversal of ΓK in ΓF . The composition cK/F ◦ rK/F is multiplication by
[K : F ].
Let K/F be an arbitrary finite field extension and M a Galois module over F . Let
E/F be the maximal separable sub-extension in K/F . As the restriction map ΓK → ΓE
∼
is an isomorphism, we have a canonical isomorphism s : H n (K, M ) → H n (E, M ). We
define the corestriction homomorphism cK/F : H n (K, M ) → H n (F, M ) as [K : E] times
the composition cE/F ◦ s.
Example 99.6. The ¡ corestriction
¢ homomorphism cK/F : H 1 (K, µm ) → H 1 (F, µm )
takes a class (x)m to NK/F (x) m .
Example 99.7. The restriction map in Galois cohomology agrees with the restriction
map for Brauer groups defined in §98.C. The corestriction in Galois cohomology yields
a map cK/F : Br(K) → Br(F ) for a finite field extension K/F . Since the composition
cK/F ◦ rK/F is the multiplication by m = [K : F ] we have Br(K/F ) ⊂ Brm (F ).
406
Let K/F be a finite separable field extension and M a Galois module over K. We
view ΓK as a subgroup of ΓF . Denote by IndK/F (M ) the group MapΓK (ΓF , M ) of ΓK -
equivariant maps ΓF → M , i.e., maps f : ΓF → M satisfying f (ρδ) = ρf (δ) for all
ρ ∈ ΓK and δ ∈ ΓF . The group IndK/F (M ) has a structure of Galois module over F
defined by (γf )(δ) = f (δγ) for all f ∈ IndK/F (M ) and γ, δ ∈ ΓF . Consider the ΓK -
module homomorphisms
u v
M− → IndK/F (M ) →
− M
defined by v(f ) = f (1) and
½
m if γ ∈ ΓK
u(m)(γ) =
0 otherwise.
Fact 99.8. Let M be a Galois module over F and K/F a finite separable field exten-
sion. Then the compositions
¡ ¢ rK/F ¡ ¢ H n (K,v)
H n F, IndK/F (M ) −−−→ H n K, IndK/F (M ) −−−−−→ H n (K, M ),
H n (K,u) ¡ ¢ cK/F ¡ ¢
H n (K, M ) −−−−−→ H n K, IndK/F (M ) −−−→ H n F, IndK/F (M )
are isomorphisms inverse to each other.
Suppose, in addition, that M is a Galois module over F . Consider the ΓF -module
homomorphisms
w t
(99.9) 0→M −
→ IndK/F (M ) −
→ M → 0.
defined by w(m)(γ) = γm and
X ¡ ¢
t(f ) = γ f (γ −1 ) ,
where the sum is taken over a left transversal of ΓK in ΓF .
Corollary 99.10. (1) The composition
H n (F,w) ¡ ¢ ∼
H n (F, M ) −−−−−→ H n F, IndK/F (M ) → H n (K, M )
coincides with rK/F .
(2) The composition
∼ ¡ ¢ H n (F,t)
H n (K, M ) → H n F, IndK/F (M ) −−−−→ H n (K, M )
coincides with cK/F .
99.D. Residue homomorphism. Let m be an integer. A Galois module M over
F is said to be m-periodic if mM = 0. If m is not divisible by char F , we write M (−1)
for the Galois module Hom(µm , M ) with the action of ΓF given by (γf )(ξ) = γf (γ −1 ξ)
for every f ∈ M (−1) (the construction is independent of the choice of m). For example,
µm (−1) = Z/mZ.
Let L be a field with a discrete valuation v and residue field F . Suppose that the
inertia group of an extension of v to Lsep acts trivially on M . Then M has a natural
structure of a Galois module over F .
100. MILNOR K-THEORY OF FIELDS 407
Fact 99.11. [48, §7] Let L be a field with a discrete valuation v and residue field F .
Let M be an m-periodic Galois module L with m not divisible by char F such that the
inertia group of an extension of v to Lsep acts trivially on M . Then there exist a residue
homomorphism ¡ ¢
∂v : H n+1 (L, M ) → H n F, M (−1)
satisfying
¡ ¢
(1) If M = µm and n = 0 then ∂v (x)m = v(x) +¡ mZ for every ¢ x ∈ L× .
(2) For every x ∈ L× with v(x) = 0, we have ∂v α ∪ (x)m = ∂v (α) ∪ (x̄)m , where
α ∈ H n+1 (L, M ) and x̄ ∈ F × is the residue of x.
Let X be a variety (integral scheme) over F and x ∈ X a regular point of codimension
1. The local ring OX,x is a discrete valuation ring with quotient field F (X) and residue
field F (x). For any m-periodic Galois module M over F let
¡ ¢ ¡ ¢
∂x : H n+1 F (X), M → H n F (x), M (−1)
denote the residue homomorphism ∂v of the associated discrete valuation v on F (X).
Fact 99.12. [48, Th. 9.2] For every field F , the sequence
r ¡ ¢ (∂x ) a n ¡ ¢ c ¡ ¢
0 → H n+1 (F, M ) −
→ H n+1 F (t), M −−→ H F (x), M (−1) → − H n F, M (−1) → 0,
x∈P1
where c is the direct sum of the corestriction homomorphisms cF (x)/F , is exact.

√
99.E. A long exact sequence. Let K = F ( a) be a quadratic field extension of a
field F of characteristic not 2. Let M be a 2-periodic Galois module over F .
We have the exact sequence (99.9) of Galois modules over F . By Corollary 99.10, the
induced exact sequence of Galois cohomology groups reads as follows
∂ rK/F cK/F ∂
− H n (F, M ) −−−→ H n (K, M ) −−−→ H n (F, M ) −
... → → H n+1 (F, M ) → . . . .
We now compute the connecting homomorphisms ∂. If n = 0 and M = Z/2Z, we have
the exact sequence
Z/2Z −→ Z/2Z →
0 ∂
− F × /F ×2 → K × /K ×2 .
The kernel of the last homomorphism is the cyclic group {1, (a)}. It follows that ∂(1 +
2Z) = (a). By Fact 99.4, the homomorphisms ∂ : H n (F, M ) → H n+1 (F, M ) coincide with
the cup-product by (a).
We have proven
√
Theorem 99.13. Let K = F ( a) be a quadratic field extension of a field F of char-
acteristic not 2 and M a 2-periodic Galois module over F . Then the following sequence
∪(a) rK/F cK/F ∪(a) rK/F
. . . −−→ H n (F, M ) −−−→ H n (K, M ) −−−→ H n (F, M ) −−→ H n+1 (F, M ) −−−→ . . .
is exact.
100. Milnor K-theory of fields

A more detailed exposition on the Milnor K-theory of fields is available in [45, Ch.
IX] and [49, Ch. 7].
408
100.A. Definition. Let F be a field. Let T denote the tensor ring of the multiplica-
tive group F × . That is a graded ring with Tn the nth tensor power of F × over Z. For
instance, T0 = Z, T1 = F × , T2 = F × ⊗Z F × etc. The graded Milnor ring K∗ (F ) of F is
the factor ring of T by the ideal generated by tensors of the form a ⊗ b with a + b = 1.
The class of a tensor a1 ⊗a2 ⊗. . .⊗an in K∗ (F ) is denoted by {a1 , a2 , . . . , an }F or simply
by {a1 , a2 , . . . , an } and is called a symbol . We have Kn (F ) = 0 if n < 0, K0 (F ) = Z,
K1 (F ) = F × . For n ≥ 2, Kn (F ) is generated (as an abelian group) by the symbols
{a1 , a2 , . . . , an } with ai ∈ F × that are subject to the following defining relations:
(M1) (Multilinearity)
{a1 , . . . , ai a0i , . . . , an } = {a1 , . . . , ai , . . . , an } + {a1 , . . . , a0i , . . . , an };
(M2) (Steinberg Relation) {a1 , a2 , . . . , an } = 0 if ai + ai+1 = 1 for some i ∈ [1, n − 1].
Note that the operation in the group Kn (F ) is written additively. In particular,
{ab} = {a} + {b} in K1 (F ) where a, b ∈ F × .
The product in the ring K∗ (F ) is given by the rule
{a1 , a2 , . . . , an } · {b1 , b2 , . . . , bm } = {a1 , a2 , . . . , an , b1 , b2 , . . . , bm }.
Fact 100.1. (1) For a permutation σ ∈ Sn , we have
{aσ(1) , aσ(2) , . . . , aσ(n) } = sgn(σ){a1 , a2 , . . . , an }
(2) {a1 , a2 , . . . , an } = 0 if ai + aj = 0 or 1 for some i 6= j.
A field homomorphism F → L induces the restriction graded ring homomorphism
rL/F : K∗ (F ) → K∗ (L) taking a symbol {a1 , a2 , . . . , an }F to {a1 , a2 , . . . , an }L . In particu-
lar, K∗ (L) has a natural structure of a left and right graded K∗ (F )-module. The image
rL/F (α) of an element α ∈ K∗ F is also denoted by αL .
If E/L is another field extension, then rE/F = rE/L ◦ rL/F . Thus, K∗ is a functor from
the category of fields to the category of graded rings.
Proposition 100.2. Let L/F be a quadratic field extension. Then
¡ ¢
Kn (L) = rL/F Kn−1 (F ) · K1 (L)
for every n ≥ 1, i.e., K∗ (L) is generated by K1 (L) as a left K∗ (F )-module.
Proof. It is sufficient to treat the ¡case n =
¢ 2. Let x, y ∈ L \ F . If x = cy for some
×
c ∈ F then {x, y} = {−c, y} ∈ rL/F K1 (F ) · K1 (L). Otherwise, as x, y and 1 are
linearly dependent over F , there are a, b ∈ F × such that ax + by = 1. We have
0 = {ax, by} = {x, y} + {x, b} + {a, by},
¡ ¢
hence {x, y} = {b}L · {x} − {a}L · {by} ∈ rL/F K1 (F ) · K1 (L). ¤
We write k∗ (F ) for the graded ring K∗ (F )/2K∗ (F ). Abusing notation, if {a1 , . . . , an }
is a symbol in Kn (F ), we shall also write it for its coset {a1 , . . . , an } + 2Kn (F ) in kn (F ).
We need some relations among symbols in k2 (F ).
Lemma 100.3. We have the following relations in k2 (F ):
(1) {a, x2 − ay 2 } = 0 for all a ∈ F × , x, y ∈ F satisfying x2 − ay 2 6= 0.
(2) {a, b} = {a + b, ab(a + b)} for all a, b ∈ F × satisfying a + b 6= 0.
Proof. (1) The statement follows from Fact 100.1 if x = 0. Suppose x 6= 0. By the
Steinberg relation, we have
0 = {a(yx−1 )2 , 1 − a(yx−1 )2 } = {a, x2 − ay 2 }.
(2) Since a(a + b) + b(a + b) is a square, by (1) we have
0 = {a(a + b), b(a + b)} = {a, b} + {a + b, ab(a + b)}. ¤
100.B. Residue homomorphism. Let L be a field with a discrete valuation v and

residue field F . The homomorphism L× → Z given by the valuation, can be viewed as
a homomorphism K1 (L) → K0 (F ). More generally, for every n ≥ 0, there is the residue
homomorphism
∂v : Kn+1 (L) → Kn (F )
uniquely determined by the following condition:
If a0 , a1 , . . . , an ∈ L× satisfying v(ai ) = 0 for all i = 1, 2, . . . n then
¡ ¢
∂v {a0 , a1 , . . . , an } = v(a0 ) · {ā1 , . . . , ān },
where ā ∈ F denotes the residue of a.
Fact 100.4. (1) If α ∈ K∗ (L) and a ∈ L× satisfies v(a) = 0 then
∂v (α · {a}) = ∂v (α) · {ā} and ∂v ({a} · α) = −{ā} · ∂v (α).
(2) Let K/L be a field extension and u a discrete valuation of K extending v with
residue field E. Let e denote the ramification index. Then for every α ∈ K∗ (L),
¡ ¢ ¡ ¢
∂u rK/L (α) = e · rE/F ∂v (α) .
100.C. Milnor’s theorem. Let X be a variety (integral scheme) over F and x ∈ X

a regular point of codimension 1. The local ring OX,x is a discrete valuation ring with
quotient field F (X) and residue field F (x). Denote by
¡ ¢ ¡ ¢
∂x : K∗+1 F (X) → K∗ F (x)
the residue homomorphism of the associated discrete valuation on F (X).

The following description of the K-groups of the function field F (t) = F (A1F ) of the
affine line is known as Milnor’s Theorem.
Fact 100.5. (Milnor’s Theorem) For every field F , the sequence

rF (t)/F ¡ ¢ (∂x ) a ¡ ¢
0 → Kn+1 (F ) −−−−→ Kn+1 F (t) −−→ Kn F (x) → 0
x∈A1(0)
is split exact.
410
100.D. Specialization. Let L be a field and v a discrete valuation on L with residue

field F . If π ∈ L× is a prime element, i.e., v(π) = 1, we define the specialization homo-
morphism
sπ : K∗ (L) → K∗ (F )
by the formula sπ (u) = ∂({−π} · u). We have
¡ ¢
sπ {a1 , a2 , . . . , an } = {b̄1 , b̄2 , . . . , b̄n },
where bi = ai /π v(ai ) .
Example 100.6. Consider the discrete valuation v of the field of rational functions
F (t) given by the irreducible polynomial t. For¡every¢ u ∈ K∗ (F ), we have st (uF (t) ) = u.
In particular, the homomorphism K∗ (F ) → K∗ F (t) is split injective as stated in Fact
100.5.
100.E. Norm homomorphism. Let L/F be a finite field extension. The standard
norm homomorphism L× → F × can be viewed as a homomorphism K1 (L) → K1 (F ). In
fact, there exists the norm homomorphism
cL/F : Kn (L) → Kn (F )
for every n ≥ 0 defined as follows.
Suppose first that the field extension L/F is simple, i.e., L is generated by one element
¡ y 1∈ ¢AF . Let α ∈
1
over F . We ¡identify
¢ L with the residue field F (y) of a closed point
Kn (L) = Kn F (y) . By Milnor’s Theorem 100.5, there is β ∈ Kn+1 F (AF ) satisfying
½
α if x = y
∂x (β) =
0 otherwise.
Let v be the discrete valuation of the field F (P1F ) = F (A1F ) associated with the infinite
point of the projective line P1F . We set cL/F (α) = ∂v (β).
In the general case, we choose a sequence of simple field extensions
F = F0 ⊂ F1 ⊂ · · · ⊂ Fn = L
and set
cL/F = cF1 /F0 ◦ cF2 /F1 ◦ · · · ◦ cFn /Fn−1 .
It turns out that the norm map cL/F is well defined, i.e., it does not depend on the
choice of the sequence of simple field extensions and the identifications with residue fields
of closed points of the affine line (cf. [45, Ch. IX, §3], [49, Ch. 7, §3]).
The following theorem is the direct consequence of the definition of the norm map and
Milnor’s Theorem 100.5.
Theorem 100.7. For every field F , the sequence
rF (t)/F ¡ ¢ (∂x ) a ¡ ¢ c
0 → Kn+1 (F ) −−−−→ Kn+1 F (t) −−→ Kn F (x) −
→ Kn (F ) → 0
x∈P1(0)
is exact where c is the direct sum of the corestriction homomorphisms cF (x)/F .

Fact 100.8. (1) (Transitivity) Let L/F and E/L be finite field extensions. Then
cE/F = cL/F ◦ cE/L .
(2) The norm map cL/F : K0 (L) → K0 (F ) is multiplication by [L : F ] on Z. The

norm map cL/F : K1 (L) → K1 (F ) is the classical norm L× → F × .
(3) (Projection Formula) Let L/F be a finite field extension. Then for every α ∈ K∗ F
and β ∈ K∗ (L) we have
cL/F (rL/F (α) · β) = α · cL/F (β),
i.e., if we view K∗ (L) as a K∗ (F )-module via rL/F then cL/F is a homomorphism
of K∗ (F )-modules. In particular, the composition cL/F ◦ rL/F is multiplication by
[L : F ].
(4) Let L/F be a finite field extension and v a discrete valuation on F . Let v1 , v2 , . . . , vs
be all the extensions of v to L. Then the following diagram is commutative:
(∂v ) `s ¡ ¢
Kn+1 (L) −−−i→ Kn L(vi )
i=1
 P
cL/F 
y
 c
y L(vi )/F (v)
v ∂ ¡ ¢
Kn+1 (F ) −−−→ Kn F (v) .
(5) Let L/F be a finite and E/F an arbitrary field extension. Let P1 , P2 , . . . , Pk be
all the prime (maximal) ideals of the ring R = L ⊗F E. For every i ∈ [1, k],
let Ri denote the residue field R/Pi and li the length of the localization ring RPi .
Then the following diagram is commutative:
(rR /L ) `k
Kn (L) −−−i−→ i=1 Kn (Ri )
 P
cL/F 
y
 li ·c
y Ri /E
rE/F
Kn (F ) −−−→ Kn (E).
We now turn to fields of positive characteristic.
Fact 100.9. [62, Th. A] Let F be a field of characteristic p > 0. Then the p-torsion
part of K∗ (F ) is trivial.
Fact 100.10. [62, Cor. 6.5] Let F be a field of characteristic p > 0. Then the natural
homomorphism
¡ ¢
Kn (F )/pKn (F ) → H 0 F, Kn (Fsep )/pKn (Fsep )
is an isomorphism.
Now consider the case of purely inseparable quadratic extensions.
Lemma 100.11. Let L/F be a purely inseparable quadratic field extension. Then the
composition rL/F ◦ cL/F on Kn (L) is the multiplication by 2.
Proof. The statement is obvious if n = 1. The general case follows from Proposition
100.2 and Fact 100.8(3). ¤
Write kn (E) := Kn (E)/2Kn (E) for a field E.

412
Proposition 100.12. Let L/F be a purely inseparable quadratic field extension. Then
the sequence
rL/F cL/F rL/F
kn (F ) −−−→ kn (L) −−→ kn (F ) −−−→ kn (L)
is exact.
Proof. Let α ∈ Kn (F ) satisfy αL = 2β for some β ∈ Kn (L). By Proposition 100.8,
2α = cL/F (αL ) = cL/F (2β) = 2cL/F (β),
hence α = cL/F (β) in view of Fact 100.9.
Let β ∈ Kn (L) satisfy cL/F (β) = 2α for some α ∈ Kn (F ). It follows from Lemma
100.11 that
2β = cL/F (β)L = 2αL ,
hence β = αL again by Fact 100.9. ¤
101. The cohomology groups H n,i (F, Z/mZ)

Let F be a field. For all n, m, i ∈ Z with m > 0, we define the group H n,i (F, Z/mZ)
as follows (cf. [48]): If m is not divisible by char F we set
¡ ¢
H n,i (F, Z/mZ) = H n F, µ⊗i
m ,
m = Hom(µm , Z/mZ) if i < 0. If

⊗i
where µm is the ith tensor power of µm if i ≥ 0 and µ⊗i ⊗−i
char F = p > 0 and m is power of p, we set


 Ki (F¡ )/mKi (F ) ¢ if n = i
H (F, Z/mZ) =
n,i 1
H F, Ki (Fsep )/mKi (Fsep ) if n = i + 1

0 otherwise.
In the general case, write m = m1 m2 , where m1 is not divisible by char F and m2 is a
power of char F if char F > 0, and set
H n,i (F, Z/mZ) = H n,i (F, Z/m1 Z) ⊕ H n,i (F, Z/m2 Z).
Note that if char F does not divide m and µm ⊂ F × , we have a natural isomorphism
H n,i (F, Z/mZ) ' H n,0 (F, Z/mZ) ⊗ µ⊗i
m.
In particular, the groups H n,i (F, Z/mZ) and H n,0 (F, Z/mZ) are (non-canonically) iso-
morphic.
Example 101.1. For an arbitrary field F , we have canonical isomorphisms
(1) H 0,0 (F, Z/mZ) ' Z/mZ,
(2) H 1,1 (F, Z/mZ) ' F × /F ×m ,
(3) H 1,0 (F, Z/mZ) ' Homc (ΓF , Z/mZ), H 1,0 (F, Z/2Z) ∼
= Ét2 (F ),
(4) H 2,1 (F, Z/mZ) ' Brm (F ).
101. THE COHOMOLOGY GROUPS H n,i (F, Z/mZ) 413
If L/F is a field extension, there is the restriction homomorphism
rL/F : H n,i (F, Z/mZ) → H n,i (L, Z/mZ).
If L is a finite over F , we define the corestriction homomorphism
cL/F : H n,i (L, Z/mZ) → H n,i (F, Z/mZ)
as follows: It is sufficient to consider the following two cases.
(i) If L/F is separable then cL/F is the corestriction homomorphism in Galois cohomology.
(ii) If L/F is purely inseparable then ΓL = ΓF , we have [Lsep : Fsep ] = [L : F ] and cL/F
is induced by the corestriction homomorphism K∗ (Lsep ) → K∗ (Fsep ).
Example 101.2. Let L/F be a finite field extension. By Example 99.6, the map
cL/F : L× /L×m = H 1,1 (L, Z/mZ) → H 1,1 (F, Z/mZ) = F × /F ×m
is induced by the norm map NL/F : L× → F × . If char F = p > 0, it follows from Example
99.1(2) that the map
cL/F : L/℘(L) = H 1,0 (L, Z/pZ) → H 1,0 (F, Z/pZ) = F/℘(F )
is induced by the trace map TrL/F : L → F .
Let l, m ∈ Z. If char F does not divide l and m, we have a natural exact sequence of
Galois modules
1 → µ⊗i ⊗i ⊗i
l → µlm → µm → 1
for every i. If l and m are powers of char F > 0 then by Fact 100.9, the sequence of Galois
modules
0 → Kn (Fsep )/lKn (Fsep ) → Kn (Fsep )/lmKn (Fsep ) → Kn (Fsep )/mKn (Fsep ) → 0
is exact. Taking the long exact sequences of Galois cohomology groups yields the following
proposition.
Proposition 101.3. For any l, m, n, i ∈ Z with l, m > 0, there is a natural long exact
sequence
· · · → H n,i (F, Z/lZ) → H n,i (F, Z/lmZ) → H n,i (F, Z/mZ) → H n+1,i (F, Z/lZ) → · · ·
The cup-product in Galois cohomology and the product in the Milnor ring induce a
structure of the graded ring on the graded abelian group
a
H ∗,∗ (F, Z/mZ) = H i,j (F, Z/mZ)
i,j∈Z
for every m ∈ Z. The product in this ring will be denoted by ∪.

414
101.A. Norm residue homomorphism. Let symbol (a)m denote the element in
H 1,1 (F, Z/mZ) corresponding to a ∈ F × under the isomorphism in Example 101.1(2).
Lemma 101.4. (Steinberg Relation) Let a, b ∈ F × satisfy a+b = 1. Then (a)m ∪(b)m =
0 in H 2,2 (F, Z/mZ).
Proof. We may assume that char F does not divide m. Let K = F [t]/(tm − a) and
α ∈ K be the class of t. We have a = αm and NK/F (1 − α) = b. It follows from the
Projection Formula and Example 99.6 that
¡ ¢
(a)m ∪ (b)m = cK/F rK/F (a)m ∪ (1 − α)m = 0
since rK/F (a)m = m(α)m = 0 in H 1,1 (K, Z/mZ). ¤
It follows from Lemma 101.4 that for every n, m ∈ Z there is a unique norm residue
homomorphism
(101.5) hn,m
F : Kn (F )/mKn (F ) → H n,n (F, Z/mZ)
taking the class of a symbol {a1 , a2 , . . . , an } to the cup-product (a1 )m ∪(a2 )m ∪· · ·∪(an )m .
The norm residue homomorphism allows us to view H ∗,∗ (F, Z/mZ) as a module over
the Milnor ring K∗ (F ).
By Example 99.1, the map hn,m F is an isomorphism for n = 0 and 1. Bloch and Kato
conjectured that hn,m
F is always an isomorphism.
For every l, m ∈ Z, we have a commutative diagram
Kn (F )/lmKn (F ) −−−→ Kn (F )/mKn (F )
 
 hn,m
hn,lm
F y y F
H n,n (F, Z/lmZ) −−−→ H n,n (F, Z/mZ)
with top map the natural surjective homomorphism.
The following important result was originally conjectured by Milnor in [109] and was
proven in [143] by Voevodsky.
Fact 101.6. Let F be a field of characteristic not two. If m is a power of 2 then the
norm residue homomorphism hn,mF is an isomorphism.
Proposition 101.3 and commutativity of the diagram above yield
Corollary 101.7. Let l and m be powers of 2. Then the natural homomorphism
H (F, Z/lmZ) → H n,n (F, Z/mZ) is surjective and the sequence
n,n
0 → H n+1,n (F, Z/lZ) → H n+1,n (F, Z/lmZ) → H n+1,n (F, Z/mZ)

is exact for any n.
Now consider the case m = 2. We shall write hnF for hn,2 n
F and H (F ) for H
n,n
(F, Z/2Z).
The norm residue homomorphisms commute with field extension homomorphisms.
They also commute with residue and corestriction homomorphisms as the following two
propositions show.
101. THE COHOMOLOGY GROUPS H n,i (F, Z/mZ) 415
Proposition 101.8. Let L be a field with a discrete valuation v and residue field F
of characteristic different from 2. Then the diagram
v∂
kn+1 (L) −−−→ kn (F )
 
n+1  hn
hL y y F
∂
H n+1 (L) −−−
v
→ H n (F )
is commutative.
Proof. Fact 99.11(1) shows that the diagram is commutative when n = 0. The
general case follows from Fact 99.11(2) as the group kn+1 (L) is generated by symbols
{a0 , a1 , . . . , an } with v(a1 ) = · · · = v(an ) = 0. ¤
Proposition 101.9. Let L/F be a finite field extension. Then the diagram
cL/F
kn (L) −−−→ kn (F )
 
 hn
Ly
hn y F
cL/F
H n (L) −−−→ H n (F )
is commutative.
Proof. We may assume that L/F is a simple field extension. The statement follows
from the definition of the norm map for the Milnor K-groups, Fact 99.12, and Proposition
101.8. ¤
√Proposition 101.10. Let F be a field of characteristic different from 2 and L =

F ( a)/F a quadratic extension with a ∈ F × . Then the following infinite sequence
{a} rL/F cL/F {a}
. . . → kn−1 (F ) −−→ kn (F ) −−−→ kn (L) −−→ kn (F ) −−→ kn+1 (F ) → . . .
is exact.
Proof. It follows from Proposition 101.9 that the diagram
{a} rL/F cL/F {a}
kn−1 (F ) −−−→ kn (F ) −−−→ k∗ (L) −−−→ kn (F ) −−−→ kn+1 (F )
    
 n n n hn+1
hn−1
F y hF y hL y hF y y F
{a} rL/F cL/F {a}
H n−1 (F ) −−−→ H n (F ) −−−→ H ∗ (L) −−−→ H n (F ) −−−→ H n+1 (F )
is commutative. By Fact 101.6, the vertical homomorphisms are isomorphisms. By The-
orem 99.13, the bottom sequence is exact. The result follows. ¤
Now consider the case char F = 2. The product in the Milnor ring and the cup-product
in Galois cohomology yield a pairing
K∗ (F ) ⊗ H ∗ (F ) → H ∗ (F )
making H ∗ (F ) a module over K∗ (F ).
416
· ¸
a, b
Example 101.11. By Example 99.5, we have {a} · [Fab ] = in Br2 (F ) for all
F
a ∈ F × and b ∈ F .
Proposition 101.12. Let F be a field of characteristic 2 and L/F a separable qua-
dratic field extension. Then the following sequence
rL/F cL/F ·[L] rL/F cL/F
0 → kn (F ) −−−→ kn (L) −−→ kn (F ) −−→ H n+1 (F ) −−−→ H n+1 (L) −−→ H n+1 (F ) → 0
is exact where the middle map is multiplication by the class of L in H 1 (F ).
Proof. We shall show that the sequence in question coincides with the exact se-
quence in Theorem 99.13 for the quadratic field extension ¡L/F and ¢ the Galois module
0
kn (Fsep ) over F .¡ Indeed,
¢ by Fact 100.10, we have H (E, k n E sep ) '¡ kn (E) and¢ by defi-
1 n+1 2
nition H (E, kn Esep ) ' H (E) for every field E. Note that H F, kn (Fsep ) = 0 by
Example 99.1(2). The connecting homomorphism in the sequence in Theorem 99.13 is
multiplication by the class of L in H 1 (F ). ¤
Now let F be a field of characteristic different from 2. The connecting homomorphism
bn : H n (F ) → H n+1 (F )
with respect to the short exact sequence
(101.13) 0 → Z/2Z → Z/4Z → Z/2Z → 0
is called the Bockstein map.
Proposition 101.14. The Bockstein map is trivial if n is even and coincides with
multiplication by (−1) if n is odd.
Proof. If n is even or −1 ∈ F ×2 then µ⊗n 4 ' Z/4Z and the statement follows from
Corollary 101.7.
Suppose
√ that n is odd and −1 ∈ / F ×2 . In this case µ⊗n
4 ' µ4 . Consider the field
K = F ( −1). By Theorem 99.13, the connecting homomorphism H n (F ) → H n+1 (F )
with respect to the exact sequence (99.9) is the cup-product with (−1). The classes of
the sequences (101.13) and (99.9) differ in Ext1Γ (Z/2Z, Z/2Z) by the class of the sequence
0 → Z/2Z → µ⊗n
4 → Z/2Z → 0.
By Corollary 101.7, the connecting homomorphism H n (F ) → H n+1 (F ) with respect to

this exact sequence is trivial. It follows that bn is the cup-product with (−1). ¤
101.B. Cohomological dimension and p-special fields. Let p be a prime integer.
A field F is called p-special if the degree of every finite field extension of F is a power of
p.
The following property of p-special fields is very useful.
Proposition 101.15. Let F be a p-special field and L/F a finite field extension. Then
there is a tower of field extensions
F = F0 ⊂ F1 ⊂ · · · ⊂ Fn−1 ⊂ Fn = L
satisfying [Fi+1 : Fi ] = p for all i ∈ [0, n − 1].
102. LENGTH AND HERBRAND INDEX 417
Proof. The result is clear if L/F is purely inseparable. So we may assume that L/F
is a separable extension. Let E/F be a normal closure of L/F . Set G = Gal(E/F ) and
H = Gal(E/L). As G is a p-group, there is a sequence of subgroups
G = H0 ⊃ H1 ⊃ · · · ⊃ Hn−1 ⊃ Hn = H
with the property [Hi : Hi+1 ] = p for all i ∈ [0, n − 1]. Then the fields Fi = LHi satisfy
the required properties. ¤
Proposition 101.16. For every prime integer p and field F , there is a field extension
L/F satisfying
(1) L is p-special.
(2) The degree of every finite sub-extension of L/F is not divisible by p.
−n
Proof. If char F = q > 0 and different from p, we set F 0 := ∪F q (in a fixed
algebraic closure of F ), otherwise F 0 := F . Let Γ be the Galois group of Fsep
0
/F 0 and
0
∆ ⊂ Γ a Sylow p-subgroup. The field of ∆-invariant elements L = (Fsep )∆ satisfies the
required conditions. ¤
We call the field L in Proposition 101.16 a p-special closure of F .
Let F be a field and let p be a prime integer. The cohomological p-dimension of F ,
denoted cdp (F ), is the smallest integer such that for every n > cdp (F ) and every finite
field extension L/F we have H n,n−1 (L, Z/pZ) = 0.
Example 101.17. (1) cdp (F ) = 0 if and only if F has no separable finite field exten-
sions of degree a power of p.
(2) cdp (F ) ≤ 1 if and only if Brp (L) = 0 for all finite field extensions L/F .
(3) If F is p-special, then cdp (F ) < n if and only if H n,n−1 (F, Z/pZ) = 0.
102. Length and Herbrand index

102.A. Length. Let A be a commutative ring and M an A-module of finite length.
The length of M is denoted by lA (M ). The ring A is artinian if the A-module M = A is
of finite length. We write l(A) for lA (A).
Lemma 102.1. Let C be a flat B-algebra where B and C are commutative local artinian
rings. Then for every finitely generated B-module M , we have
lC (M ⊗B C) = l(C/QC) · lB (M ),
where Q is the maximal ideal of B.
Proof. Since C is flat over B, both sides of the equality are additive in M . Thus, we
may assume that M is a simple B-module, i.e., M = B/Q. We have M ⊗B C ' C/QC
and the equality follows. ¤
Setting M = B we obtain
Corollary 102.2. In the conditions of Lemma 102.1, one has l(C) = l(C/QC)·l(B).
418
Lemma 102.3. Let B be a commutative A-algebra and M a B-module of finite length

over A. Then X
lA (M ) = lBQ (MQ ) · lA (B/Q),
where the sum is taken over all maximal ideals Q ⊂ B.
Proof. Both sides are additive in M , so we may assume that M = B/Q, where Q is
a maximal ideal of B. The result follows. ¤
102.B. Herbrand index. Let M be a module over a commutative ring A and a ∈ A.
a
Suppose that the modules M/aM and a M := ker(M −
→ M ) have finite length. The integer
h(a, M ) = lA (M/aM ) − lA (a M )
is called the Herbrand index of M relative to a.
We collect simple properties of the Herbrand index in the following lemma.
Lemma 102.4. (1) Let 0 → M 0 → M → M 00 → 0 be an exact sequence of A-
modules. Then h(a, M ) = h(a, M 0 ) + h(a, M 00 ).
(2) If M has finite length then h(a, M ) = 0.
Lemma 102.5. Let S be a one-dimensional Noetherian local ring and P1 , . . . , Pm all
the minimal prime ideals of S. Let M be a finitely generated S-module and s ∈ S not
belonging to any of Pi . Then
m
X ¡ ¢
h(s, M ) = lSPi (MPi ) · l S/(Pi + sS) .
i=1
Proof. Since s ∈
/ Pi , the coset of s in S/Pi is not a zero divisor. Hence
¡ ¢
lS S/(Pi + sS) = h(s, S/Pi ).
Both sides of the equality are additive in M . Since M has a filtration with factors S/P ,
where P is a prime ideal of S, we may assume that M = S/P . If P is maximal then
MPi = 0 and h(s, M ) = 0 since M is simple. If P = Pi for some i then lSPj (M ) = 1 if
i = j and zero otherwise. The equality holds in this case too. ¤
103. Places
Let K be a field. A valuation ring R of K is a subring R ⊂ K such that for any
x ∈ K \ R, we have x−1 ∈ R. A valuation ring is a local domain. A trivial example of a
valuation ring is the field K itself.
Given two fields K and L, a place π : K * L is a local ring homomorphism f : R → L
of a valuation ring R ⊂ K. We say that the place π is defined on R. An embedding of
fields is a trivial example of a place defined everywhere. A place K * L is called surjective
if f is surjective.
If K and L are extensions of a field F , we say that a place K * L is an F -place if π
is defined and the identity on F .
Let K * L and L * E be two places, given by ring homomorphisms f : R → L
and g : S → E respectively, where R ⊂ K and S ⊂ L are valuation rings. Then the
f |T g
ring T = f −1 (S) is a valuation ring of K and the composition T −−→ S →
− E defines
103. PLACES 419
the composition place K * E. In particular, any place L * E can be restricted to any

subfield K ⊂ L.
A composition of two F -places is an F -place. Every place is a composition of a
surjective place and a field embedding.
A place K * L is said to be geometric, if it is a composition of (finitely many) places
each defined on a discrete valuation ring. An embedding of fields will be also viewed as a
geometric place.
Let Y be a complete variety over F and let π : F (Y ) * L be an F -place. The valuation
ring R of the place dominates a unique point y ∈ Y , i.e., OY,y ⊂ R and the maximal
ideal of OY,y is contained in the maximal ideal M of R. The induced homomorphism
of fields F (y) → R/M → L over F gives rise to an L-point of Y , i.e., to a morphism
f : Spec(L) → Y with image {y}. We say that y is the center of π and f is induced by π.
Let X be a regular variety over F and let f : Spec(L) → X be a morphism over F .
Choose a regular system of parameters a1 , a2 , . . . , an in the local ring R = OX,x , where {x}
is the image of f . Let Mi be the ideal of R generated by a1 , . . . , ai and set Ri = R/Mi ,
Pi = Mi+1 /Mi . Denote by Fi the quotient field of Ri , in particular, F0 = F (X) and
Fn = F (x). The localization ring (Ri )Pi is a discrete valuation ring with quotient field
Fi and residue field Fi+1 , therefore it determines a place Fi * Fi+1 . The composition of
places
F (X) = F0 * F1 * . . . * Fn = F (x) ,→ L
is a geometric place constructed (non-canonically) out of the point f ∈ X(L).
Lemma 103.1. Let K be an arbitrary field, K 0 /K an odd degree field extension, and
L/K an arbitrary field extension. Then there exists a field L0 , containing K 0 and L, such
that the extension L0 /L is of odd degree.
Proof. We may assume that K 0 /K is a simple extension, i.e., K 0 is generated over
K by one element. Let f (t) ∈ F [t] be its minimal polynomial. Since the degree of f
is odd, there exists an irreducible divisor g ∈ L[t] of f over L with odd deg(g). We set
L0 = L[t]/(g). ¤
Lemma 103.2. Let K be a field extension of F of finite transcendence degree over F ,
K * L a geometric F -place and K 0 a finite field extension of K of odd degree. Then
there exists an odd degree field extension L0 /L such that the place K * L extends to a
place K 0 * L0 .
Proof. By Lemma 103.1, it suffices to consider the case of a surjective place K * L
given by a discrete valuation ring R. It is also suffices to consider only two cases: (1)
K 0 /K is purely inseparable and (2) K 0 /K is separable.
In the first case, the degree [K 0 : K] is a power of an odd prime p. Let R0 be an
arbitrary valuation ring of K 0 lying over R, i.e., such that R0 ∩ K = R and with the
embedding R → R0 local (such an R0 exists in the case of an arbitrary field extension
K 0 /K by [147, Ch. VI Th. 50 ]). We have a surjective place K 0 * L0 , where L0 is the
residue field of R0 . We claim that L0 is purely inseparable over L (and therefore [L0 : L],
n
being a power of p, is odd). Indeed if l ∈ L0 , choose a preimage k ∈ R0 of l. Then k p ∈ K
n
for some n hence lp ∈ L, i.e., L0 /L is a purely inseparable extension.
420
In the second case, consider all the valuation rings R1 , . . . , Rr of K 0 lying over R (the
number of such valuation rings is finite by [147, Ch. P VI, Th. 12, Cor. 4]). The residue
field of each Ri is a finite extension of L. Moreover, ri=1 ei ni = [K 0 : K] by [147, Ch.
VI, Th. 20 and p. 63], where ni is the degree over L of the residue field of Ri , and ei is
the ramification index of Ri over R (cf. [147, Def. on pp. 52–53]). It follows that at least
one of the ni is odd. ¤
104. Cones and vector bundles

The word “scheme” in the next two sections means a separated scheme of finite type
over a field.
104.A. Definition of a cone. Let X be a scheme over a field F and let S • =
S ⊕ S 1 ⊕ S 2 ⊕ . . . be a sheaf of graded OX -algebras. We assume that
0
(1) the natural morphism OX → S 0 is an isomorphism;

(2) the OX -module S 1 is coherent;
(3) the sheaf of OX -algebras S • is generated by S 1 .
The cone of S • is the scheme C = Spec(S • ) over X and P(C) = Proj(S • ) is called the
projective cone of C. Recall that Proj(S • ) has a covering by the principal open subschemes
D(s) := Spec(S(s) ) over all s ∈ S 1 , where S(s) is the subring of degree 0 elements in the
localization Ss .
We have natural morphisms C → X and P(C) → X. The canonical homomorphism
S • → S 0 of OX -algebras induces the zero section X → C.
If C and C 0 are cones over X, then C ×X C 0 has a natural structure of a cone over X.
We denote it by C ⊕ C 0 .
¡ ¢
Example 104.1. A coherent OX -module P defines the cone C(P ) = Spec S • (P )
over X, where S • stands for the symmetric algebra. If the sheaf P is locally free, the
cone E := C(P ) is called the vector bundle over X with the dual sheaf of sections P ∨ :=
HomOX (P, OX ). The projective cone P(E) is called the projective bundle of E. The
assignment P 7→ C(P ∨ ) gives rise to an equivalence between the category of locally free
coherent OX -modules and the category of vector bundles over X. In particular, such
operations over the locally free OX -modules as the tensor product, symmetric power,
dual sheaf etc., and the notion of an exact sequence translate to the category of vector
bundles. We write K0 (X) for the Grothendieck group of the category of vector bundles
over X. The group K0 (X) is the abelian group given by generators the isomorphism
classes [E] of vector bundles E over X and relations [E] = [E 0 ] + [E 00 ] for every exact
sequence 0 → E 0 → E → E 00 → 0 of vector bundles over X.
Example 104.2. The trivial line bundle X × A1 → X will be denoted by 1.
Example 104.3. Let f : Y → X be a closed embedding and I ⊂ OX the sheaf of
ideals of the image of f in X. The cone
¡ ¢
Cf = Spec OX /I ⊕ I/I 2 ⊕ I 2 /I 3 ⊕ . . .
over Y is called the normal cone of Y in X. If X is a scheme of pure dimension d then
Cf is also a scheme of pure dimension d [47, B.6.6].
104. CONES AND VECTOR BUNDLES 421
Example 104.4. If f : X → C is the zero section of a cone C then Cf = C.

Example 104.5. The cone TX := Cf of the diagonal embedding f : X → X × X is
called the tangent cone of X. If X is a scheme of pure dimension d then the tangent cone
TX is a scheme of pure dimension 2d (cf. Example 104.3).
Let U and V be vector spaces over a field F and let
U = U0 ⊃ U1 ⊃ U2 ⊃ . . . and V = V0 ⊃ V1 ⊃ V2 ⊃ . . .
be two filtrations by subspaces. Consider the filtration on U ⊗ V defined by
X
(U ⊗ V )k = Ui ⊗ Vj .
i+j=k
The following lemma can be proven by a suitable choice of bases of U and V .

Lemma 104.6. The canonical linear map
a
(Ui /Ui+1 ) ⊗ (Vj /Vj+1 ) → (U ⊗ V )k /(U ⊗ V )k+1
i+j=k
is an isomorphism for every k ≥ 0.

Proposition 104.7. Let f : Y → X and g : S → T be closed embeddings. Then there
is a canonical isomorphism Cf × Cg ' Cf ×g .
Proof. We may assume that X = Spec(A), Y = Spec(A/I) and T = Spec(B),
S = Spec(B/J), where I ⊂ A and J ⊂ B are ideals. Then X × T = Spec(A ⊗ B) and
Y × S = Spec(A ⊗ B)/K, where K = I ⊗ B + A ⊗ J.
Consider the vector spaces Ui = I i and Vj = J j . We have (U ⊗ V )k = K k . By Lemma
104.6,
¡a i i+1 a j j+1 ¢ ¡a k k+1 ¢
Cf × Cg = Spec I /I ⊗ J /J ' Spec K /K = Cf ×g . ¤
i≥0 j≥0 k≥0
Corollary 104.8. If X and Y are two schemes then TX×Y = TX × TY .

104.B. Regular closed embeddings. Let A be a commutative ring. A sequence
a = (a1 , a2 , . . . , ad ) of elements of A is called regular if the coset of ai is not a zero divisor
in the factor ring A/(a1 A + · · · + ai−1 A) for all i ∈ [1, d]. We write l(a) = d.
Let Y be a scheme and d : Y → Z a locally constant function. A closed embedding
f : Y → X is called regular of codimension d is for every point y ∈ Y there is an affine
neighborhood U ⊂ X of f (y) such that the ideal of f (Y ) ∩ U in F [U ] is generated by a
regular sequence of length d(y).
Let f : Y → X be a ` closed embedding and I the sheaf of ideals of Y in OX . The
embedding of I/I 2 into k≥0 I k /I k+1 induces a surjective OY -algebra homomorphism
`
S • (I/I 2 ) → k≥0 I k /I k+1 and therefore a closed embedding of cones ϕf : Cf → C(I/I 2 )
over Y .
Proposition 104.9 ([50, Cor. 16.9.4, Cor. 16.9.11]). A closed embedding f : Y → X
is regular of codimension d if and only if the OY -module I/I 2 is locally free of rank d and
the natural morphism ϕf : Cf → C(I/I 2 ) is an isomorphism.
422
Corollary 104.10. Let f : Y → X be a regular closed embedding of codimension d

and I the sheaf of ideals of Y in OX . Then the normal cone Cf is a vector bundle over
Y of rank d with the sheaf of sections naturally isomorphic to (I/I 2 )∨ .
We shall write Nf for the normal cone Cf of a regular closed embedding f and call
Nf the normal bundle of f .
Proposition 104.11. Let f : Y → X be a closed embedding and g : X 0 → X a
faithfully flat morphism. Then f is a regular closed embedding if and only if the closed
embedding f 0 : Y 0 = Y ×X X 0 → X 0 is regular.
Proof. Let I be the sheaf of ideals of Y in OX . Then I 0 = g ∗ (I) is the sheaf of ideals
of Y 0 in OX 0 . Moreover
k k+1 2
g ∗ (I k /I k+1 ) = I 0 /I 0 , Cf ×Y Y 0 = Cf 0 , C(I/I 2 ) ×Y Y 0 = C(I 0 /I 0 )
and ϕf ×Y 1Y 0 = ϕf 0 . By faithfully flat descent, I/I 2 is locally free and ϕf is an isomor-
phism if and only if I 0 /I 0 2 is locally free and ϕf 0 is an isomorphism. The statement follows
by Proposition 104.9. ¤
Proposition 104.12 ([50, Cor. 17.12.3]). Let g : X → Y be a smooth morphism of
relative dimension d and f : Y → X a section of g, i.e., g ◦ f = 1Y . Then f is a regular
closed embedding of codimension d and Nf = f ∗ Tg , where Tg := ker(TX → g ∗ TY ) is the
relative tangent bundle of g over X.
Corollary 104.13. Let X be a smooth scheme. Then the diagonal embedding X →
X × X is regular. In particular, the tangent cone TX is a vector bundle over X.
Proof. The diagonal embedding is a section of any of the two projections X × X →
X. ¤
If X is a smooth scheme, the vector bundle TX is called the tangent bundle over X.
Corollary 104.14. Let f : X → Y be a morphism where Y is a smooth scheme of
pure dimension d. Then the morphism h = (1X , f ) : X → X × Y is a regular closed
embedding of codimension d with Nh = f ∗ TY .
Proof. Applying Proposition 104.12 to the smooth projection p : X × Y → X of
relative dimension d, we have the closed embedding h is regular of codimension d. The
tangent bundle Tp is equal to q ∗ TY , where q : X × Y → Y is the other projection. Since
q ◦ h = f , we have
Nh = h∗ Tp = h∗ ◦ q ∗ TY = f ∗ TY . ¤
Proposition 104.15 ([50, Prop. 19.1.5]). Let g : Z → Y and f : Y → X be regular
closed embeddings of codimension s and r respectively. Then f ◦ g is a regular closed
embedding of codimension r + s and the natural sequence of normal bundles over Z
0 → Ng → Nf ◦g → g ∗ Nf → 0
is exact.
Proposition 104.16 ([50, Th. 17.12.1, Prop. 17.13.2]). A closed embedding f : Y →

X of smooth schemes is regular and the natural sequence of vector bundles over Y
0 → T Y → f ∗ T X → Nf → 0
is exact.
104.C. Canonical line bundle. Let C = Spec(S • ) be a cone over X. The cone
Spec(S • [t]) = C × A1 coincides with C ⊕ 1. Let I ⊂ S • [t] be the ideal generated by S 1 . The
closed subscheme of P(C ⊕ 1) defined by I is isomorphic to Proj(S 0 [t]) = Spec(S 0 ) = X.
Thus we get a canonical closed embedding (canonical section) of X into P(C ⊕ 1).
Set Lc := P(C ⊕ 1) \ X. The inclusion of S(s) •
into S • [t](s) for every s ∈ S 1 induces a
morphism Lc → P(C).
Proposition 104.17. The morphism Lc → P(C) has a canonical structure of a line
bundle.
• t
Proof. We have S • [t](s) = S(s) [ s ], hence the preimage of D(s) is isomorphic to D(s)×
A . For any other element s ∈ S we have st0 = ss0 st , i.e., the change of coordinate function
1 0 1
is linear. ¤
The line bundle Lc → P(C) is called the canonical line bundle over P(C).
•
A section of Lc over the open set D(s) is given by an S(s) -algebra homomorphism
• t • t
S(s) [ s ] → S(s) that is uniquely determined by the image as of s . The element sas ∈ Ss of
degree 1 agrees with s0 as0 on the intersection D(s) ∩ D(s0 ). Therefore the sheaf of sections
of Lc coincides with S f• (1) = O(1).
The scheme P(C) can be viewed as a locally principal divisor of P(C ⊕ 1) given by t.
The open complement P(C ⊕ 1) \ P(C) is canonically isomorphic to C. The image of the
canonical section X → P(C ⊕ 1) belongs to C (and in fact is equal to the¡ image of ¢the
zero section of C), hence it does not intersect P(C). Moreover, P(C ⊕ 1) \ P(C) ∪ X is
canonically isomorphic to C \ X.
If C is a cone over X, we write C ◦ for C \ X where X is viewed as a closed subscheme
of C via the zero section. We have shown that C ◦ is canonically isomorphic to L◦c . Note
that C is a cone over X and Lc is a cone (in fact, a line bundle) over P(C).
For every s ∈ S 1 , the localization Ss is the Laurent polynomial ring S(s) [s, s−1 ] over
S(s) . Hence the inclusion of S(s) into Ss induces a flat morphism C ◦ → P(C) of relative
dimension 1.
104.D. Tautological line bundle. Let C = Spec(S • ) be a cone over X. Consider

the tensor product T • = S • ⊗S 0 S • as a graded ring with respect to the second factor. We
have
Proj(T • ) = C ×X P(C).
Let J be the ideal of T • generated by x ⊗ y − y ⊗ x for all x, y ∈ S 1 and set
¡ ¢
Lt := Proj T • /J .
Thus Lt is a closed subscheme of C ×X P(C) and we have natural projections Lt → C
and Lt → P(C).
424
Proposition 104.18. The morphism Lt → P(C) has a canonical structure of a line

bundle.
¡ • ¢
Proof. Let s ∈ S 1 . The preimage of D(s) in Lt coincides with Spec T(1⊗s) /J(1⊗s) ,
•
where J(1⊗s) = J1⊗s ∩T(1⊗s) . The homomorphism T • → Ss• [t], where t is a variable, defined
by x ⊗ y 7→ xy sn
· tn for any x ∈ S n and y ∈ S • , gives rise to an isomorphism between
•
T(1⊗s) •
/J(1⊗s) and S(s) [t]. Hence the preimage of D(s) is isomorphic to D(s) × A1 . ¤
The line bundle Lt → P(C) is called the tautological line bundle over P(C).
Example 104.19. If L is a line bundle over X, then P(L) = X and Lt = L ×X P(L) =
L.
Similar to the case of the canonical line bundle, a section of Lt over the open set D(s)
•
is given by an element as ∈ S(s) and the element as /s ∈ Ss• of degree −1 agrees with
as0 /s0 on the intersection D(s) ∩ D(s0 ). Therefore the sheaf of section of Lt coincides with
f• (−1) = O(−1). In particular, the tautological line bundle is dual to the canonical line
S
bundle, Lt = Lc ∨ .
The ideal I = S >0 in S • defines the image of the zero section of C. The graded ring
T • /J is isomorphic to S • ⊕ I ⊕ I 2 ⊕ · · · . Therefore, the canonical morphism Lt → C is
the blow up of C along the image of the zero section of C. The exceptional divisor in
Lt is the image of the zero section of Lt . Hence the induced morphism L◦t → C ◦ is an
isomorphism.
Example 104.20. Let F [ε] be the F -algebra of dual number over F . The tangent
space TP(V ),L of the point of the projective space P(V ) given by a line L ⊂ V coincides with
the fiber over L of the map P(V )(F [ε]) → P(V )(F ) induced by the ring homomorphism
F [ε] → F , ε 7→ 0. For example, the F [ε]-submodule L ⊕ Lε of V [ε] := V ⊗ F [ε] represents
the zero vector of the tangent space TP(V ),L .
For a linear map h : L → V let Wh be the F [ε]-submodule of V [ε] generated by the
elements v + h(v)ε, v ∈ L. Since Wh /εWh ' L, we can view Wh as a point of TP(V ),L . The
map HomF (L, V ) → TP(V ),L given by h 7→ Wh yields an exact sequence of vector spaces
0 → HomF (L, L) → HomF (L, V ) → TP(V ),L → 0.
In other words,
TP(V ),L = HomF (L, V /L).
Since the fiber of the tautological line bundle Lt over the point given by L coincides
with L, we get an exact sequence of vector bundles over P(V ):
0 → Hom(Lt , Lt ) → Hom(Lt , 1 ⊗F V ) → TP(V ) → 0.
The first term of the sequence is isomorphic to 1 and the second term to Lc ⊗F V ' (Lc )⊕n ,
where n = dim V . It follows that
¡ ¢
[TP(V ) ] = n[Lc ] − 1 ∈ K0 P(V ) .
More generally, if E → X is a vector bundle then there is an exact sequence of vector
bundles over P(E):
0 → 1 → Lc ⊗ q ∗ E → Tq → 0,
where q : P(E) → X is the natural morphism and Tq is the relative tangent bundle of q.
104.E. Deformation to the normal cone. Let f : Y → X be a closed embedding

of schemes. First suppose first that X is an affine scheme, X = Spec(A), and Y is given
by an ideal I ⊂ A. Set Y = Spec(A/I). Consider the subring
a
e=
A I −n tn
n∈Z
of the Laurent polynomial ring A[t, t−1 ], where the negative powers of the ideal I are
understood as equal to A. The scheme Df := Spec(A) e is called the deformation scheme
of the closed embedding f . In the general case, in order to define Df , we cover X by open
affine subschemes and glue together the deformation schemes of the restrictions of f to
the open sets of the covering.
e induces a morphism g : Df → A1 × X. Denote by Cf the
The inclusion of A[t] into A
−1
inverse image g ({0} × X). In the affine case,
Cf = Spec(A/I ⊕ I/I 2 · t−1 ⊕ I 2 /I 3 · t−2 ⊕ . . . ).
Thus, Cf is the normal cone of f (cf. Example 104.3). If f is a regular closed embedding
of codimension d then Cf is a vector bundle over Y of rank d. We write Nf for Cf in this
case.
The open complement Df \ Cf is the inverse image g −1 (Gm × X). In the affine case,
e −1 ] = A[t, t−1 ]. Hence the inverse image is canonically
it is the spectrum of the ring A[t
isomorphic to Gm × X via g, i.e.,
Df \ Cf ' Gm × X.
In the affine case, the natural ring homomorphism A[t] → (A/I)[t] extends canonically
to a ring homomorphism A e → (A/I)[t]. Hence the morphism f × id : A1 × Y → A1 × X
factors through the canonical morphism h : A1 × Y → Df over A1 . The fiber of h over a
point t 6= 0 of A1 is naturally isomorphic to the morphism f . The fiber of h over t = 0 is
isomorphic to the zero section Y → Cf of the normal cone Cf of f . Thus we can view h
as a family of closed embeddings parameterized by A1 deforming the closed embedding f
into the zero section Y → Cf as the parameter t “approaches 0”. We have the following
diagram of natural morphisms:
Y −−−→ A1 × Y ←−−− Gm × Y
  
  
y y y
Cf −−−→ Df ←−−− Gm × X
  °
  °
y y °
Y −−−→ A1 × X ←−−− Gm × X.
Note that the normal cone Cf is the principal divisor in Df of the function t.
426
Consider a fiber product diagram

f0
Y 0 −−−→ X0
 
(104.21) 
gy

yh
f
Y −−−→ X
where f and f 0 are closed embeddings. It induces the fiber product diagram of open and
closed embeddings
Cf 0 −−−→ Df 0 ←−−− Gm × X 0
  
  id×h
(104.22) ky ly y
Cf −−−→ Df ←−−− Gm × X.
Proposition 104.23. In the notation of (104.21), there are natural closed embeddings
Df 0 → Df ×X X 0 and Cf 0 → Cf ×X X 0 . These embeddings are isomorphisms if h is flat.
Proof. We may assume that all schemes are affine and h is given by a ring ho-
momorphism A → A0 . The scheme Y is defined by an ideal I ⊂ A and Y 0 is given
by I 0 = IA0 ⊂ A0 . The natural homomorphism I n ⊗A A0 → (I 0 )n is surjective, hence
Ae ⊗A A0 → A e0 is surjective. Consequently, Df 0 → Df ×X X 0 and Cf 0 → Cf ×X X 0
are closed embeddings. If A0 is flat over A, the homomorphism I n ⊗A A0 → (I 0 )n is an
isomorphism. ¤
104.F. Double deformation space. Let A be a commutative ring.

Lemma 104.24. Let I be the ideal of A generated by a regular sequence a = (a1 , a2 , . . . , ad )
and a ∈ A satisfy a + I is not a zero divisor in A/I. If ax ∈ I m for some x ∈ A and m
then x ∈ I m .
Proof. By Proposition 104.9, multiplication by a + I on I n /I n+1 is injective for any
n. The statement of the lemma follows by induction on m. ¤
Let a = (a1 , a2 , . . . , ad ) and b = (b1 , b2 , . . . , be ) be two sequences of elements of A. We

write a ⊂ b if d ≤ e and ai = bi for all i ∈ [1, d]. Clearly, if a ⊂ b and b is regular so is a.
Let I ⊂ J be ideals of A. We define the ideals I n J m for n < 0 or m < 0 by
½ n+m
n m J if n < 0
I J =
In if m < 0.
Proposition 104.25. Let a ⊂ b be two regular sequences in a ring A and I ⊂ J the
ideals of A generated by a and b respectively. Then
I n J m ∩ I n+1 = I n+1 J m−1
I n J m ∩ J n+m+1 = I n J m+1
for all n and m.
Proof. We prove the first equality. The proof of the second one is similar.
We proceed by induction on m. The case m ≤ 1 is clear. Suppose m ≥ 2. As the
inclusion “⊃” is easy, we need to prove that
I n J m ∩ I n+1 ⊂ I n+1 J m−1 .
Let d be a sequence such that a ⊂ d ⊂ b and let L be the ideal generated by d, so
I ⊂ L ⊂ J. By descending induction on the length l(d) of the sequence d, we prove that
(104.26) I n J m ∩ I n+1 ⊂ Ln+1 J m−1 .
When l(d) = l(a), i.e., d = a and L = I, we get the desired inclusion.
The case l(d) = l(b), i.e., d = b and L = J is obvious. Let c be the sequence satisfying
a ⊂ c ⊂ d and l(c) = l(d) − 1. Let K be the ideal generated by c. We have L = K + aA
where a is the last element of the sequence d. Assuming (104.26), we shall prove that
I n J m ∩ I n+1 ⊂ K n+1 J m−1 .
Let x ∈ I n J m ∩ I n+1 . By assumption,
n+1
X
n+1 m−1
x∈L J = an+1−k K k J m−1 ,
k=0
hence
n+1
X
x= an+1−k xk
k=0
k m−1
for some xk ∈ K J . For any s ∈ [0, n + 1], set
Xs
ys = as−k xk .
k=0
s m−1
We claim that ys ∈ K J for s ∈ [0, n + 1]. We prove the claim by induction on
s. The case s = 0 is obvious since y0 = x0 ∈ J m−1 . Suppose ys ∈ K s J m−1 for some
s < n + 1. We have
Xn+1
n+1−s
x=a ys + an+1−k xk ,
k=s+1
k m−1 s+1
where xk ∈ K J ⊂K if k ≥ s + 1 and x ∈ I n+1 ⊂ K s+1 . Hence an+1−s ys ∈ K s+1
and therefore ys ∈ K s+1 by Lemma 104.24. Thus ys ∈ K s J m−1 ∩ K s+1 . By the first
induction hypothesis, the latter ideal is equal to K s+1 J m−2 and ys ∈ K s+1 J m−2 . Since
xs+1 ∈ K s+1 J m−1 , we have ys+1 = ays + xs+1 ∈ K s+1 J m−1 . This proves the claim. By
the claim, x = yn+1 ∈ K n+1 J m−1 . ¤
g f
Let Z −
→Y −
→ X be regular closed embeddings. We have closed embeddings
i : (Nf )|Z → Nf and j : Ng → Nf g .
We shall construct the double deformation scheme D = Df,g and a morphism D → A2
satisfying all of the following:
(1) D|A1 ×Gm = Df × Gm .
(2) D|Gm ×A1 = Gm × Df g .
428
(3) D|A1 ×{0} = Dj .

(4) D|{0}×A1 = Di .
(5) D|{0}×{0} = Ni ' Nj .
As in the case of an ordinary deformation space, it suffices to consider the affine case:
X = Spec(A), Y = Spec(A/I), and Z = Spec(A/J), where I ⊂ J are the ideals of A
generated by regular sequences. Consider the subring
a
b=
A I n J m−n · t−n s−m
n,m∈Z
of the Laurent polynomial ring A[t, s, t−1 , s−1 ] and set D = Spec(A). b Since A
b contains
the polynomial ring A[t, s], there are natural morphisms D → X × A → A . 2 2
We have
a ³ a ´
b −1 ] =
A[s I n · t−n s−m = I n · t−n [s, s−1 ],
n,m∈Z n,m∈Z
a ³ a ´
b −1 ] =
A[t J m · t−n s−m = J m · s−m [t, t−1 ].
n,m∈Z n,m∈Z
This proves (1) and (2).

To prove (3) consider the rings
a £ ¤
b A
A/s b= I n J m−n /I n J m−n+1 · t−n ,
n,m∈Z
a£ ¤
R= J m /J m+1 · s−m ,
m∈Z
a£ ¤
S= (J m + I)/(J m+1 + I) · s−m .
m∈Z
We have Spec(R) = Nf g and Spec(S) = Ng . The natural surjection R → S corre-

sponds to the embedding j : Ng → Nf g .
Let Ie = ker(R → S). By Proposition 104.25, J m ∩ I = IJ m−1 , hence
a£ ¤
Ie = IJ m−1 + J m+1 /J m+1 · s−m
m∈Z
and a£ ¤
Ien = I n J m−n + J m+1 /J m+1 · s−m .
m∈Z
Therefore, Dj is the spectrum of the ring
a£ ¤
I n J m−n + J m+1 /J m+1 · t−n s−m .
m∈Z
It follows from Proposition 104.25 that this ring coincides with A/sb A,
b hence (3).
To prove (4) consider the ring
a £ ¤
b A
A/t b= I n J m−n /I n+1 J m−n−1 · s−m .
n,m∈Z
105. GROUP ACTIONS ON ALGEBRAIC SCHEMES 429
The normal bundle Nf is the spectrum of the ring

a£ ¤
T = I n /I n+1 · u−m .
n∈Z
Let Je be the ideal of T of the closed subscheme (Nf )|Z . We have

a£ ¤
Jem = I n J m + I n+1 /I n+1 · u−m .
n∈Z
The deformation scheme Di is the spectrum of the ring

a £ ¤
U= I n J m + I n+1 /I n+1 · u−n s−m .
n,m∈Z
b A
We define the surjective ring homomorphism ϕ : A/t b → U taking
(x + I n+1 J m−n−1 ) · t−n s−m to (x + I n+1 ) · u−n s−m+n .

By Proposition 104.25, the map ϕ is also injective. Hence ϕ gives the identification (4).
Property (5) follows from (3) and (4).
105. Group actions on algebraic schemes

In this section all schemes are quasi-projective over a field F . We write G = {1, σ}
for a cyclic group of order 2.
105.A. G-schemes. A G-scheme is a scheme Y together with a G-action on Y . As Y

is a quasi-projective scheme, every pair of points of Y belong to an open affine subscheme.
It follows that there is an open G-invariant affine covering of such an Y . Therefore, in
most of the constructions and proofs, we may restrict to the class of affine G-schemes.
In particular, to define a subscheme Y G ⊂ Y for a G-scheme Y , we may assume that
Y is affine, i.e., Y = Spec R, where R is a G-algebra, i.e., a commutative F -algebra with
G acting on R by F -algebra automorphisms. Consider the ideal I ⊂ R generated by
σ(r) − r for all r ∈ R. Set
Y G := Spec(R/I).
Example 105.1. For any scheme X over F , the group G acts on X ×X by permutation
of the factors. Then (X × X)G coincides with the image of the diagonal closed embedding
∆ : X → X × X. Indeed, if X = Spec(A) then X × X = Spec(A ⊗F A) and the ideal
I ⊂ A ⊗F A is generated by σ(a ⊗ a0 ) − a ⊗ a0 = a0 ⊗ a − a ⊗ a0 for all a, a0 ∈ A, and hence
coincides with the kernel of the product map A ⊗ A → A.
Exercise 105.2. Prove that (X × Y )G = X G × Y G for G-schemes X and Y .
Proposition 105.3. Let Y be G-scheme and B the blow up of Y along Y G . Then
G acts naturally on B and the subscheme B G of B coincides with the exceptional divisor
P(C), where C is the normal cone of the closed embedding Y G → Y .
430
Proof. We may assume that Y = Spec R for a G-algebra R. Then Y G = Spec(R/I),

where I is the ideal of R generated by σ(r) − r for all r ∈ R, and B = Proj S, where
S = R ⊕I ⊕I 2 ⊕· · · . The scheme B is covered by open G-invariant subschemes Spec(S(s) )
with s = σ(r) − r, r ∈ R. The intersection of Spec(S(s) ) with the exceptional divisor P(C)
is given by the ideal sS(s) in S(s) . Therefore, it suffices to show that the ideal J in S(s)
generated by elements of the form σ(x) − x with x ∈ S(s) coincides with sS(s) . Clearly
s = σ(r) − r ∈ J. Conversely, let x = srn with r ∈ I n be an element in S(s) . As σ acts on
I n /I n+1 by multiplication by (−1)n , we have t := (−1)n σ(r) − r ∈ I n+1 and hence
t t
σ(x) − x = n
= s n+1 ∈ sS(s) . ¤
s s
Let Y be G-scheme. To define a scheme Y /G we may assume that Y = Spec R. Set
Y /G := Spec(RG ).
The natural morphism p : Y → Y /G is called a G-torsor if Y G = ∅.
Example 105.4. Let Y be a G-scheme and Y 0 = Y \ Y G . Then the natural morphism
Y 0 → Y 0 /G is a G-torsor.
Proposition 105.5. Let p : Y → Y /G be a G-torsor. Then
(1) p∗ (OY ) is a locally free OY /G -module of rank 2.
(2) The morphism p is étale.
Proof. We may assume that Y = Spec(R) for a G-algebra R. As Y G = ∅, the
elements s = σ(r) − r with r ∈ R generate the unit ideal of R, i.e., the G-invariant
open sets Spec(Rs ) cover Y . Replacing R by Rs we may assume that s is invertible in
R. Then
¡ {1, r} is a¢ basis of R over RG¡ since for¢ every a ∈ R we have a = f + gr with
f = aσ(r) − σ(a)r s−1 ∈ RG and g = σ(a) − a s−1 ∈ RG .
Moreover, r is the root of the quadratic polynomial h(t) = t2 − bt + c over RG with
b = σ(r) + r and c = σ(r)r. As the element h0 (r) = 2r − b = −s is invertible in R, the
morphism p is étale (cf. [107, Ch. I, Ex. 3.4]). ¤
Example 105.6. Let p : Y → Y /G be a G-torsor with Y /G = Spec F . Then Y =
Spec K, where K is an étale quadratic F -algebra.
Proposition 105.7. Suppose char F 6= 2. Let Y be a G-scheme and U = Y \ Y G .
Then
p
(1) The composition Y G → Y − → Y /G is a closed embedding with the complement
U/G.
(2) If I ⊂ OY /G is the sheaf of ideals of Y G in Y /G, then p∗ (I) = J 2 , where J ⊂ OY
is the sheaf of ideals of Y G in Y .
(3) If Y is regular and Y G is a regular divisor in Y then Y /G is also regular.
Proof. We may assume that Y = Spec R for a G-algebra R.
(1): The composition Y G → Y → Y /G is given by the composition of algebra homomor-
phisms RG → R → R/I, where I is the ideal of R generated by¡ σ(r) − r¢ for all r ∈ R.
The later composition is surjective as any r ∈ R is the image of σ(r) + r /2 ∈ RG .
105. GROUP ACTIONS ON ALGEBRAIC SCHEMES 431
The scheme U (respectively U/G) is covered by the open sets Spec(Rs ) (respectively
Spec(RsG2 )) with s = σ(r) − r, r ∈ R. Therefore, the second statement follows from the
equality (Rs )G = (RG )s2 .
¡ ¢
(2): We need to prove that I ∩ RG R = I 2 . The generators s = σ(r) − r of I satisfy
σ(s) = −s. Therefore, the ideal I 2 is generated by G-invariant P elements. Hence I 2 ⊂
G G
(I ∩ R )R. Conversely, let a ∈ I ∩ R . We can write a = ri si with ri ∈ R and si ∈ I
satisfying σ(si ) = −si . Hence
1¡ ¢ 1 X¡ ¢
a = a + σ(a) = ri − σ(ri ) si ∈ I 2 .
2 2
(3): Let z ¡∈ Y /G and
¢ y ∈ Y a point above z. Suppose y ∈ Y \Y G . The natural morphism
Y \Y G → Y \Y G /G is a G-torsor by Example 105.4 hence is flat (cf. Proposition 105.5).
As the local ring OY,y is regular, so is OY /G,z by [50, Prop. 17.3.3(i)].
Suppose now y ∈ Y G . We may assume that F is algebraically closed and y is a
rational point. The tangent space TY G ,y has codimension 1 in TY,y and coincides with
the subspace of all G-invariant elements in TY,y . Hence there is a basis ā1 , ā2 , . . . , ān of
mY,y /m2Y,y satisfying σ(ā1 ) = −ā1 and σ(āi ) = āi for i ≥ 2. Lift the āi to a local system of
parameters ai ∈ mY,y with σ(a1 ) = −a1 and σ(ai ) = ai for i ≥ 2. The completion of the
local ring OY /G,z coincides with
¡ ¢
ObY /G,z = O bY,y G = F [[a1 , a2 , . . . , an ]]G = F [[a2 , a2 , . . . , an ]],
1
and hence is regular. Therefore, z is a regular point on Y /G. ¤

105.B. The scheme BX . Let X be a scheme. Write BX for the blow up of X 2 ×A1 :=
X × X × A1 along ∆(X) × {0}. Since the normal ¡ cone of¢ ∆(X) × {0} in X × A is TX ⊕ 1
2 1
(cf. Proposition 104.7), the projective cone P TX ⊕ 1 is the exceptional divisor in BX

(cf. [47, B.6.6]).
Let G act on X 2 × A1 by σ(x, x0 , t) = (x0 , x, −t). If char F 6= 2, we have
(X 2 × A1 )G = (X 2 )G × (A1 )G = ∆(X) × {0}.
¡ ¢
Set UX = (X 2 × A1 ) \ ∆(X)
¡ × {0}
¢ . By Proposition
¡ ¢105.3, the group G acts naturally
on BX so that (BX )G = P TX ⊕ 1 and BX \ P TX ⊕ 1 is canonically isomorphic to UX .
Proposition 105.8. Suppose char F 6= 2. If X is smooth then so are BX and BX /G.
Proof. The scheme¡ BX is ¢smooth as the blow up of a smooth scheme along smooth
center. As the scheme P TX ⊕ 1 = (BX )G is a smooth divisor in BX , the second statement
follows from Proposition 105.7(3). ¤
Bibliography
[1] A. Adrian Albert, Structure of algebras, Revised printing. American Mathematical Society Col-
loquium Publications, Vol. XXIV, American Mathematical Society, Providence, R.I., 1961. MR
MR0123587 (23 #A912)
[2] M Amer, Quadratische Formen über Funktionenkörpern., (1976), thesis.
[3] J. Arason and R. Baeza, Relations in I n and I n Wq in characteristic 2, preprint.
[4] Jón K. Arason and Albrecht Pfister, Zur Theorie der quadratischen Formen über formalreellen
Körpern, Math. Z. 153 (1977), no. 3, 289–296. MR MR0432541 (55 #5529)
[5] Jón Kr. Arason, On Krüskemper’s theorem, unpublished.
[6] , Cohomologische Invarianten quadratischer Formen, J. Algebra 36 (1975), no. 3, 448–491.
MR MR0389761 (52 #10592)
[7] , A proof of Merkurjev’s theorem, Quadratic and Hermitian forms (Hamilton, Ont., 1983),
CMS Conf. Proc., vol. 4, Amer. Math. Soc., Providence, RI, 1984, pp. 121–130. MR MR776449
(86f:11029)
[8] Jón Kr. Arason and Richard Elman, Nilpotence in the Witt ring, Amer. J. Math. 113 (1991), no. 5,
861–875. MR MR1129295 (92k:11043)
[9] , Powers of the fundamental ideal in the Witt ring, J. Algebra 239 (2001), no. 1, 150–160.
MR MR1827878 (2002b:11054)
[10] Jón Kr. Arason and Manfred Knebusch, Über die Grade quadratischer Formen, Math. Ann. 234
(1978), no. 2, 167–192. MR MR0506027 (58 #21933)
[11] Jón Kristinn Arason and Albrecht Pfister, Beweis des Krullschen Durchschnittsatzes für den Wit-
tring, Invent. Math. 12 (1971), 173–176. MR MR0294251 (45 #3320)
[12] , Beweis des Krullschen Durchschnittsatzes für den Wittring, Invent. Math. 12 (1971), 173–
176. MR MR0294251 (45 #3320)
[13] R. Aravire and R. Baeza, The behavior of quadratic and differential forms under function field
extensions in characteristic two, J. Algebra 259 (2003), no. 2, 361–414. MR MR1955526
(2003k:11060)
[14] , Annihilators of quadratic and bilinear forms over fields of characteristic two, J. Algebra
299 (2006), no. 1, 294–308. MR MR2225777
[15] Roberto Aravire and Ricardo Baeza, Milnor’s k-theory and quadratic forms over fields of charac-
teristic two, Comm. Algebra 20 (1992), no. 4, 1087–1107. MR MR1154404 (93a:11030)
[16] R. Baeza, Comparing u-invariants of fields of characteristic 2, Bol. Soc. Brasil. Mat. 13 (1982),
no. 1, 105–114. MR MR692281 (84f:10027)
[17] Ricardo Baeza, Quadratic forms over semilocal rings, Springer-Verlag, Berlin, 1978, Lecture Notes
in Mathematics, Vol. 655. MR MR0491773 (58 #10972)
[18] G. Berhuy and Z. Reichstein, On the notion of canonical dimension for algebraic groups, Adv.
Math. 198 (2005), no. 1, 128–171. MR MR2183253 (2006j:11053)
[19] Nicolas Bourbaki, Elements of mathematics. Commutative algebra, Hermann, Paris, 1972, Trans-
lated from the French. MR MR0360549 (50 #12997)
[20] , Éléments de mathématique, Masson, Paris, 1985, Algèbre commutative. Chapitres 5 à 7.
[21] Ludwig Bröcker, Zur Theorie der quadratischen Formen über formal reellen Körpern, Math. Ann.
210 (1974), 233–256. MR MR0354549 (50 #7027)
433
434 BIBLIOGRAPHY
[22] Patrick Brosnan, Steenrod operations in Chow theory, Trans. Amer. Math. Soc. 355 (2003), no. 5,
1869–1903 (electronic). MR 1 953 530
[23] Armand Brumer, Remarques sur les couples de formes quadratiques, C. R. Acad. Sci. Paris Sér.
A-B 286 (1978), no. 16, A679–A681. MR MR0498374 (58 #16502)
[24] J. W. S. Cassels, On the representation of rational functions as sums of squares, Acta Arith. 9
(1964), 79–82. MR MR0162791 (29 #95)
[25] J. W. S. Cassels, W. J. Ellison, and A. Pfister, On sums of squares and on elliptic curves over
function fields, J. Number Theory 3 (1971), 125–149. MR MR0292781 (45 #1863)
[26] J. W. S. Cassels and A. Fröhlich (eds.), Algebraic number theory, London, Academic Press Inc.
[Harcourt Brace Jovanovich Publishers], 1986, Reprint of the 1967 original. MR MR911121
(88h:11073)
[27] Vladimir Chernousov, Stefan Gille, and Alexander Merkurjev, Motivic decomposition of isotropic
projective homogeneous varieties, Duke Math. J. 126 (2005), no. 1, 137–159. MR MR2110630
(2005i:20070)
[28] Craig M. Cordes, The Witt group and the equivalence of fields with respect to quadratic forms, J.
Algebra 26 (1973), 400–421. MR MR0325656 (48 #4003)
[29] Antoine Delzant, Définition des classes de Stiefel-Whitney d’un module quadratique sur un corps
de caractéristique différente de 2, C. R. Acad. Sci. Paris 255 (1962), 1366–1368. MR MR0142606
(26 #175)
[30] M. Dickmann and F. Miraglia, Lam’s conjecture, Algebra Colloq. 10 (2003), no. 2, 149–176. MR
MR1980436 (2004e:11037)
[31] M. A. Dickmann and F. Miraglia, On quadratic forms whose total signature is zero mod 2n , Invent.
Math. 133 (1998), no. 2, 243–278. MR MR1632786 (99f:11052)
[32] R. Elman and H.-K. Lee, On a relative u-invariant, preprint.
[33] Richard Elman, Pfister forms and K-theory of fields, (1972), thesis.
[34] , Quadratic forms and the u-invariant. III, Conference on Quadratic Forms—1976 (Proc.
Conf., Queen’s Univ., Kingston, Ont., 1976), Queen’s Univ., Kingston, Ont., 1977, pp. 422–444.
Queen’s Papers in Pure and Appl. Math., No. 46. MR MR0491490 (58 #10732)
[35] Richard Elman and T. Y. Lam, Pfister forms and K-theory of fields, J. Algebra 23 (1972), 181–213.
MR MR0302739 (46 #1882)
[36] , Quadratic forms over formally real fields and pythagorean fields, Amer. J. Math. 94 (1972),
1155–1194. MR MR0314878 (47 #3427)
[37] , Quadratic forms and the u-invariant. II, Invent. Math. 21 (1973), 125–137. MR
MR0417053 (54 #5114)
[38] , Classification theorems for quadratic forms over fields, Comment. Math. Helv. 49 (1974),
373–381. MR MR0351997 (50 #4485)
[39] , Quadratic forms under algebraic extensions, Math. Ann. 219 (1976), no. 1, 21–42. MR
MR0401649 (53 #5476)
[40] Richard Elman, T. Y. Lam, and Adrian R. Wadsworth, Amenable fields and Pfister extensions,
Conference on Quadratic Forms—1976 (Proc. Conf., Queen’s Univ., Kingston, Ont., 1976), Queen’s
Univ., Kingston, Ont., 1977, pp. 445–492. With an appendix “Excellence of F (ϕ)/F for 2–fold
Pfister forms” by J. K. Arason. Queen’s Papers in Pure and Appl. Math., No. 46. MR MR0560497
(58 #27756)
[41] Richard Elman and Tsit Yuen Lam, Quadratic forms and the u-invariant. I, Math. Z. 131 (1973),
283–304. MR MR0323716 (48 #2072)
[42] Richard Elman, Tsit Yuen Lam, and Alexander Prestel, On some Hasse principles over formally
real fields, Math. Z. 134 (1973), 291–301. MR MR0330045 (48 #8384)
[43] Richard Elman and Christopher Lum, On the cohomological 2-dimension of fields, Comm. Algebra
27 (1999), no. 2, 615–620. MR MR1671946 (99m:12007)
[44] Richard Elman and Alexander Prestel, Reduced stability of the Witt ring of a field and its
Pythagorean closure, Amer. J. Math. 106 (1984), no. 5, 1237–1260. MR MR761585 (86f:11030)
BIBLIOGRAPHY 435
[45] I. B. Fesenko and S. V. Vostokov, Local fields and their extensions, second ed., Translations of
Mathematical Monographs, vol. 121, American Mathematical Society, Providence, RI, 2002, With
a foreword by I. R. Shafarevich. MR 2003c:11150
[46] Robert W. Fitzgerald, Function fields of quadratic forms, Math. Z. 178 (1981), no. 1, 63–76. MR
MR627094 (83b:10021)
[47] W. Fulton, Intersection theory, Springer-Verlag, Berlin, 1984.
[48] R. Garibaldi, A. Merkurjev, and Serre J.-P., Cohomological invariants in galois cohomology, Amer-
ican Mathematical Society, Providence, RI, 2003.
[49] P. Gille and T. Szamuely, Central simple algebras and galois cohomology, Cambridge, UK, 2006.
[50] A. Grothendieck, Éléments de géométrie algébrique. IV. Étude locale des schémas et des morphismes
de schémas IV, Inst. Hautes Études Sci. Publ. Math. (1967), no. 32, 361. MR MR0238860 (39 #220)
[51] D. K. Harrison, Witt rings, (1970).
[52] Robin Hartshorne, Algebraic geometry, Springer-Verlag, New York, 1977, Graduate Texts in Math-
ematics, No. 52. MR 57 #3116
[53] I. N. Herstein, Noncommutative rings, Mathematical Association of America, Washington, DC,
1994, Reprint of the 1968 original, With an afterword by Lance W. Small. MR 97m:16001
[54] Detlev W. Hoffmann, Isotropy of quadratic forms over the function field of a quadric, Math. Z. 220
(1995), no. 3, 461–476. MR MR1362256 (96k:11041)
[55] , On Elman and Lam’s filtration of the u-invariant, J. Reine Angew. Math. 495 (1998),
175–186. MR MR1603861 (99d:11037)
[56] , On the dimensions of anisotropic quadratic forms in I 4 , Invent. Math. 131 (1998), no. 1,
185–198. MR MR1489898 (98i:11022)
[57] , Pythagoras numbers of fields, J. Amer. Math. Soc. 12 (1999), no. 3, 839–848. MR
MR1670858 (2000d:11050)
[58] Detlev W. Hoffmann and Ahmed Laghribi, Quadratic forms and Pfister neighbors in character-
istic 2, Trans. Amer. Math. Soc. 356 (2004), no. 10, 4019–4053 (electronic). MR MR2058517
(2005e:11041)
[59] , Isotropy of quadratic forms over the function field of a quadric in characteristic 2, J.
Algebra 295 (2006), no. 2, 362–386. MR MR2194958 (2007a:11052)
[60] , Isotropy of quadratic forms over the function field of a quadric in characteristic 2, J.
Algebra 295 (2006), no. 2, 362–386. MR MR2194958 (2007a:11052)
[61] Jurgen Hurrelbrink, Nikita A. Karpenko, and Ulf Rehmann, The minimal height of quadratic forms
of given dimension, Arch. Math. 87 (2006), 522–529.
[62] O. Izhboldin, On p-torsion in K∗M for fields of characteristic p, Algebraic K-theory, Amer. Math.
Soc., Providence, RI, 1991, pp. 129–144. MR 92f:11165
[63] O. T. Izhboldin, On the cohomology groups of the field of rational functions, Mathematics in St.
Petersburg, Amer. Math. Soc., Providence, RI, 1996, pp. 21–44. MR 97e:12008
[64] Oleg T. Izhboldin, Motivic equivalence of quadratic forms, Doc. Math. 3 (1998), 341–351 (elec-
tronic). MR MR1668530 (99m:11036)
[65] , Motivic equivalence of quadratic forms. II, Manuscripta Math. 102 (2000), no. 1, 41–52.
MR MR1771227 (2001g:11051)
[66] , Fields of u-invariant 9, Ann. of Math. (2) 154 (2001), no. 3, 529–587. MR MR1884616
(2002m:11026)
[67] Bill Jacob and Markus Rost, Degree four cohomological invariants for quadratic forms, Invent.
Math. 96 (1989), no. 3, 551–570. MR MR996554 (90g:11044)
[68] B. Kahn, Formes quadratiques et cycles algébriques, Séminaire Bourbaki, no. 941, 2004-05.
[69] Irving Kaplansky, Quadratic forms, J. Math. Soc. Japan 5 (1953), 200–207. MR MR0059260
(15,500a)
[70] N. A. Karpenko, Algebro-geometric invariants of quadratic forms, Algebra i Analiz 2 (1990), no. 1,
141–162. MR 91g:11032a
[71] , Cohomology of relative cellular spaces and of isotropic flag varieties, Algebra i Analiz 12
(2000), no. 1, 3–69. MR 2001c:14076
436 BIBLIOGRAPHY
[72] Nikita Karpenko and Alexander Merkurjev, Essential dimension of quadrics, Invent. Math. 153
(2003), no. 2, 361–372. MR MR1992016 (2004f:11029)
[73] Nikita A. Karpenko, On the first Witt index of quadratic forms, Invent. Math. 153 (2003), no. 2,
455–462. MR MR1992018 (2004f:11030)
[74] , Holes in I n , Ann. Sci. École Norm. Sup. (4) 37 (2004), no. 6, 973–1002. MR MR2119244
(2005k:11082)
[75] Nikita A. Karpenko and Alexander S. Merkurjev, Canonical p-dimension of algebraic groups, Adv.
Math. 205 (2006), no. 2, 410–433. MR MR2258262
[76] Kazuya Kato, Symmetric bilinear forms, quadratic forms and Milnor K-theory in characteristic
two, Invent. Math. 66 (1982), no. 3, 493–510. MR MR662605 (83i:10027)
[77] , Symmetric bilinear forms, quadratic forms and Milnor K-theory in characteristic two,
Invent. Math. 66 (1982), no. 3, 493–510. MR 83i:10027
[78] , Milnor K-theory and the Chow group of zero cycles, Applications of algebraic K-theory to
algebraic geometry and number theory, Part I, II (Boulder, Colo., 1983), Contemp. Math., vol. 55,
Amer. Math. Soc., Providence, RI, 1986, pp. 241–253. MR MR862638 (88c:14012)
[79] Yoshiyuki Kitaoka, Arithmetic of quadratic forms, Cambridge Tracts in Mathematics, vol. 106,
Cambridge University Press, Cambridge, 1993. MR MR1245266 (95c:11044)
[80] M. Knebusch, Generic splitting of quadratic forms. I, Proc. London Math. Soc. (3) 33 (1976), no. 1,
65–93.
[81] Manfred Knebusch, Ein Satz über die Werte von quadratischen Formen über Körpern, Invent.
Math. 12 (1971), 300–303. MR MR0289546 (44 #6734)
[82] , Specialization of quadratic and symmetric bilinear forms, and a norm theorem, Acta Arith.
24 (1973), 279–299, Collection of articles dedicated to Carl Ludwig Siegel on the occasion of his
seventy-fifth birthday, III. MR MR0349582 (50 #2075)
[83] , Generic splitting of quadratic forms. I, Proc. London Math. Soc. (3) 33 (1976), no. 1,
65–93. MR MR0412101 (54 #230)
[84] , Generic splitting of quadratic forms. II, Proc. London Math. Soc. (3) 34 (1977), no. 1,
1–31. MR MR0427345 (55 #379)
[85] Manfred Knebusch and Winfried Scharlau, Algebraic theory of quadratic forms, DMV Seminar,
vol. 1, Birkhäuser Boston, Mass., 1980, Generic methods and Pfister forms, Notes taken by Heisook
Lee. MR MR583195 (82b:10023)
[86] M.-A. Knus, A. Merkurjev, M. Rost, and J.-P. Tignol, The book of involutions, American Mathe-
matical Society, Providence, RI, 1998, With a preface in French by J. Tits.
[87] Max-Albert Knus, Alexander Merkurjev, Markus Rost, and Jean-Pierre Tignol, The book of in-
volutions, American Mathematical Society Colloquium Publications, vol. 44, American Mathe-
matical Society, Providence, RI, 1998, With a preface in French by J. Tits. MR MR1632779
(2000a:16031)
[88] Martin Krüskemper, On real local-global principles, Math. Z. 204 (1990), no. 1, 145–151. MR
MR1048071 (91f:11028)
[89] T. Y. Lam, Ten lectures on quadratic forms over fields, Conference on Quadratic Forms—1976
(Proc. Conf., Queen’s Univ., Kingston, Ont., 1976), Queen’s Univ., Kingston, Ont., 1977, pp. 1–
102. Queen’s Papers in Pure and Appl. Math., No. 46. MR MR0498380 (58 #16508)
[90] , Introduction to quadratic forms over fields, Graduate Studies in Mathematics, vol. 67,
American Mathematical Society, Providence, RI, 2005. MR MR2104929 (2005h:11075)
[91] Serge Lang, Algebra, third ed., Graduate Texts in Mathematics, vol. 211, Springer-Verlag, New
York, 2002. MR MR1878556 (2003e:00003)
[92] H.-K. Lee, Relative u-invariants of algebraic quadratic forms, thesis (1995).
[93] David B. Leep, The Amer-Brumer theorem over arbitrary fields, (1971), preprint.
[94] , Systems of quadratic forms, J. Reine Angew. Math. 350 (1984), 109–116. MR MR743536
(85j:11038)
[95] D. W. Lewis, New proofs of the structure theorems for Witt rings, Exposition. Math. 7 (1989),
no. 1, 83–88. MR MR982158 (90d:11050)
BIBLIOGRAPHY 437
[96] F. Lorenz and J. Leicht, Die Primideale des Wittschen Ringes, Invent. Math. 10 (1970), 82–88.
MR MR0266949 (42 #1851)
[97] P. Mammone, R. Moresi, and A. R. Wadsworth, u-invariants of fields of characteristic 2, Math. Z.
208 (1991), no. 3, 335–347. MR MR1134579 (92m:11033)
[98] Murray Marshall, A reduced theory of quadratic forms, Conference on Quadratic Forms—1976
(Proc. Conf., Queen’s Univ., Kingston, Ont., 1976), Queen’s Univ., Kingston, Ont., 1977, pp. 569–
579. Queen’s Papers in Pure and Appl. Math., No. 46. MR MR0491491 (58 #10733)
[99] Hideya Matsumoto, Sur les sous-groupes arithmétiques des groupes semi-simples déployés, Ann.
Sci. École Norm. Sup. (4) 2 (1969), 1–62. MR MR0240214 (39 #1566)
[100] Hideyuki Matsumura, Commutative algebra, second ed., Mathematics Lecture Note Series, vol. 56,
Benjamin/Cummings Publishing Co., Inc., Reading, Mass., 1980.
[101] A. S. Merkur0 ev, On the norm residue symbol of degree 2, Dokl. Akad. Nauk SSSR 261 (1981),
no. 3, 542–547. MR MR638926 (83h:12015)
[102] , On the norm residue symbol of degree 2, Dokl. Akad. Nauk SSSR 261 (1981), no. 3,
542–547. MR 83h:12015
[103] , Kaplansky’s conjecture in the theory of quadratic forms, Zap. Nauchn. Sem. Leningrad.
Otdel. Mat. Inst. Steklov. (LOMI) 175 (1989), no. Koltsa i Moduli. 3, 75–89, 163–164. MR
MR1047239 (90m:11058)
[104] , Simple algebras and quadratic forms, Izv. Akad. Nauk SSSR Ser. Mat. 55 (1991), no. 1,
218–224. MR 93b:16025
[105] A. S. Merkur0 ev and A. A. Suslin, The group K3 for a field, Izv. Akad. Nauk SSSR Ser. Mat. 54
(1990), no. 3, 522–545. MR MR1072694 (91g:19002)
[106] A. Meyer, über die auflösung der gleichung ax2 + by 2 + cz 2 + du2 + ev 2 = 0 in ganzen zahlen,
Vierteljahrs. Naturfors. Gesells 29 (1884), 220–222.
[107] J. S. Milne, Étale cohomology, Princeton University Press, Princeton, N.J., 1980. MR 81j:14002
[108] John Milnor, Algebraic K-theory and quadratic forms, Invent. Math. 9 (1969/1970), 318–344. MR
MR0260844 (41 #5465)
[109] , Algebraic K-theory and quadratic forms, Invent. Math. 9 (1969/1970), 318–344. MR 41
#5465
[110] D. Orlov, A. Vishik, and V. Voevodsky, An exact sequence for K∗M /2 with applications to quadratic
forms, Ann. of Math. (2) 165 (2007), no. 1, 1–13.
[111] Albrecht Pfister, Multiplikative quadratische Formen, Arch. Math. 16 (1965), 363–370. MR
MR0184937 (32 #2408)
[112] , Zur Darstellung von −1 als Summe von Quadraten in einem Körper, J. London Math.
Soc. 40 (1965), 159–165. MR MR0175893 (31 #169)
[113] , Quadratische Formen in beliebigen Körpern, Invent. Math. 1 (1966), 116–132. MR
MR0200270 (34 #169)
[114] , Quadratische Formen in beliebigen Körpern, Invent. Math. 1 (1966), 116–132. MR
MR0200270 (34 #169)
[115] , Quadratic forms with applications to algebraic geometry and topology, London Mathemat-
ical Society Lecture Note Series, vol. 217, Cambridge University Press, Cambridge, 1995. MR
MR1366652 (97c:11046)
[116] Alexander Prestel, Remarks on the Pythagoras and Hasse number of real fields, J. Reine Angew.
Math. 303/304 (1978), 284–294. MR MR514686 (80d:12020)
[117] D. Quillen, Higher algebraic K-theory. I, (1973), 85–147. Lecture Notes in Math., Vol. 341. MR 49
#2895
[118] Ulf Rehmann and Ulrich Stuhler, On K2 of finite-dimensional division algebras over arithmetical
fields, Invent. Math. 50 (1978/79), no. 1, 75–90. MR 80g:12012
[119] M. Rost, Chow groups with coefficients, Doc. Math. 1 (1996), No. 16, 319–393 (electronic). MR
98a:14006
[120] Markus Rost, Hilbert theorem 90 for K3M for degree-two extensions, University of Regensburg
(1986), preprint.
438 BIBLIOGRAPHY
[121] , On quadratic forms isotropic over the function field of a conic, Math. Ann. 288 (1990),
no. 3, 511–513. MR MR1079875 (91i:11044)
[122] Chih Han Sah, Symmetric bilinear forms and quadratic forms, J. Algebra 20 (1972), 144–160. MR
MR0294378 (45 #3448)
[123] Winfried Scharlau, Zur Pfisterschen Theorie der quadratischen Formen, Invent. Math. 6 (1969),
327–328. MR MR0241453 (39 #2793)
[124] , Quadratic reciprocity laws, J. Number Theory 4 (1972), 78–97. MR MR0292753 (45 #1835)
[125] , Quadratic and Hermitian forms, Grundlehren der Mathematischen Wissenschaften [Fun-
damental Principles of Mathematical Sciences], vol. 270, Springer-Verlag, Berlin, 1985. MR
MR770063 (86k:11022)
[126] C. Schubert, Going up the u-invariant over formally real fields, Math. Ann. 338 (2007), no. 3,
757–765.
[127] J.-P. Serre, A course in arithmetic, Springer-Verlag, New York, 1973, Translated from the French,
Graduate Texts in Mathematics, No. 7. MR MR0344216 (49 #8956)
[128] Jean-Pierre Serre, Galois cohomology, english ed., Springer Monographs in Mathematics, Springer-
Verlag, Berlin, 2002, Translated from the French by Patrick Ion and revised by the author. MR
MR1867431 (2002i:12004)
[129] , Galois cohomology, english ed., Springer-Verlag, Berlin, 2002, Translated from the French
by Patrick Ion and revised by the author. MR 1 867 431
[130] Igor R. Shafarevich, Basic algebraic geometry. 1, second ed., Springer-Verlag, Berlin, 1994, Varieties
in projective space, Translated from the 1988 Russian edition and with notes by Miles Reid. MR
MR1328833 (95m:14001)
[131] T. A. Springer, Quadratic forms over fields with a discrete valuation. I. Equivalence classes of
definite forms, Nederl. Akad. Wetensch. Proc. Ser. A. 58 = Indag. Math. 17 (1955), 352–362. MR
MR0070664 (17,17a)
[132] Tonny Albert Springer, Sur les formes quadratiques d’indice zéro, C. R. Acad. Sci. Paris 234 (1952),
1517–1519. MR MR0047021 (13,815j)
[133] A. A. Suslin, Quaternion homomorphism for the field of functions on a conic, Dokl. Akad. Nauk
SSSR 265 (1982), no. 2, 292–296. MR 84a:12030
[134] , Torsion in K2 of fields, K-Theory 1 (1987), no. 1, 5–29. MR MR899915 (89a:11123)
[135] Richard G. Swan, Zero cycles on quadric hypersurfaces, Proc. Amer. Math. Soc. 107 (1989), no. 1,
43–46. MR 89m:14005
[136] Jean-Pierre Tignol, Réduction de l’indice d’une algèbre simple centrale sur le corps des fonctions
d’une quadrique, Bull. Soc. Math. Belg. Sér. A 42 (1990), no. 3, 735–745, Algebra, groups and
geometry. MR MR1316221 (95k:16021)
[137] B. Totaro, Birational geometry of quadrics in characteristic 2, Linear Algebraic Groups and Related
Structures, http://www.math.uni-bielefeld.de/lag/ (2006).
[138] Burt Totaro, Birational geometry of quadrics in characteristic 2, Linear Algebraic Groups and
Related Structures (preprint server) 225 (2006), 1–15.
[139] A. Vishik, On the Chow groups of quadratic Grassmannians, Doc. Math. 10 (2005), 111–130 (elec-
tronic). MR MR2148072 (2006b:14007)
[140] Alexander Vishik, Motives of quadrics with applications to the theory of quadratic forms, Geometric
methods in the algebraic theory of quadratic forms, Lecture Notes in Math., vol. 1835, Springer,
Berlin, 2004, pp. 25–101. MR MR2066515 (2005f:11055)
[141] , Fields of u-invariant 2r + 1, (2006), preprint.
[142] Vladimir Voevodsky, Motivic cohomology with Z/2-coefficients, Publ. Math. Inst. Hautes Études
Sci. (2003), no. 98, 59–104. MR MR2031199 (2005b:14038b)
[143] , Motivic cohomology with Z/2-coefficients, Publ. Math. Inst. Hautes Études Sci. (2003),
no. 98, 59–104. MR MR2031199
[144] , Reduced power operations in motivic cohomology, Publ. Math. Inst. Hautes Études Sci.
(2003), no. 98, 1–57. MR MR2031198
BIBLIOGRAPHY 439
[145] Adrian R. Wadsworth, Merkurjev’s elementary proof of Merkurjev’s theorem, Applications of al-
gebraic K-theory to algebraic geometry and number theory, Part I, II (Boulder, Colo., 1983),
Contemp. Math., vol. 55, Amer. Math. Soc., Providence, RI, 1986, pp. 741–776. MR MR862663
(88b:11078)
[146] Ernst Witt, Theorie der quadratischen Formen in beliebigen Körpern, J. Reine Angew. Math. 176
(1937), 31–44.
[147] O. Zariski and P. Samuel, Commutative algebra. Vol. II, Springer-Verlag, New York, 1975, Reprint
of the 1960 edition, Graduate Texts in Mathematics, Vol. 29. MR 52 #10706
440 BIBLIOGRAPHY
Index
1-fold Pfister form ū-invariant, 161

quadratic, 48 and finite extensions, 163, 165
An , see Property An and quadratic extensions, 164
G-algebra, 427 ūr (F )-invariant, 162
G-scheme, 427 p-special closure, 415
G-torsor, 428 u-invariant, 161
In and finite extensions, 170
and torsion-freeness, 183 and quadratic extensions, 167, 170
exact sequences and quadratic exten- u0 -invariant, 166
sions, 178, 179 Che(X 2 ), 321
n
I (F ), 17 dimIzh , 331
and Jn (F ), 180 mult, 333
fundamental ideal I(F ), 17
generators absolute Galois group, 401
¡ √ of, ¢17 absolute stability index, 191
if I n F ( −1) = 0, 159
maps additive completion, 300
affine bundle, 245
ē0 , 17
algebra
ē1 , 18
étale, 394
ē1 is an isomorphism, 18
quadratic, 395
ē2 is an isomorphism, 22
quaternion, 397
e0 , 17
semisimple, 394
e1 , 17
separable, 394
Fact: ēn is an isomorphism, 64
split, 394
presentation of, 186
central, 395
Presentation of I 2 (F ), 20
opposite, 396
torsion in, 183 split, 395
torsion-freeness going down, 190 algebras
2
Iq (F ) Brauer equivalent, 395
maps Arf invariant, 58
ē2 is an isomorphism in character- Artin-Schreier map, 35, 395
istic two, 64 associativity law, 304
n
Iq (F )
maps basic monomial, 368
ē1 is an isomorphism, 58 bilinear form, 3
en , 63 group of similarity factors, 7
en well-defined on Pfister forms, 63 alternating, 3
J-invariant, 372 anisotropic, 9
Jn (F ), 104 anisotropic part, 11
and I n (F ), 180 determinant, 4
K-cohomology groups, 260 diagonal, 7
K-homology group, 242 dimension, 3
INDEX 441
isotropic, 9 Chow group of dimension p cycles, 265

lagrangian, 9 Chow motive, 300
metabolic, 9 Classification Theorem, 151
non-degenerate, 3, 75 Clifford algebra, 52
radical, 5 even, 53
round, 7 clopen set, 391
signature of, 123 cocycle relation, 19
signed determinant, 17 cohomological p-dimension, 415
skew-symmetric, 3 cohomology group, 401
subform, 4 cone, 418
support of, 135 normal, 418
symmetric, 3 projective, 418
tensor product, 12 tangent, 419
totally isotropic subspace, 9 connecting homomorphism, 243
universal, 8 corestriction homomorphism, 403, 411
Witt index, 11 correspondence, 295, 299, 332
Bilinear Similarity Norm Principle, 83 incidence, 366
Bilinear Similarity Theorem, 84 multiplicity of, 333
Bilinear Substitution Principle, 68 prime, 332
Bilinear Value Norm Principle, 75 cup-product, 402
Bilinear Value Theorem, 73 cycle
Bockstein map, 414 q-primordial, 328
boolean space, 391 basic, 318
boundary map, 232 derivative of, 321
Brauer group, 395 essence of, 318
essential, 318
canonical integer produced by 1-primordial, 329
p-dimension, 375 minimal, 349
dimension, 375 primordial, 325, 349
involution, 397 rational, 317, 370
line bundle, 421 subcycle of, 318
Cartan Formula, 293 cycle of a subscheme, 221
Cassels-Pfister Theorem, 67
category deformation homomorphism, 238, 243
of Chow motives, 303 deformation scheme, 423
of correspondences, 300 degree
of graded Chow motives, 303 homomorphism, 267
of graded correspondences, 300 of a cycle, 315
chain equivalence of a subvariety, 315
bilinear forms, 14 of an algebra, 396
Chain Lemma, 399 derivative of a cycle, 321
change of field homomorphism, 225 diagonalizable bilinear form, 7
character group, 401 dimension map, 16
Chern class, 251 divisor, 221
Chow group of codimension p, 271 double deformation scheme, 425
442 BIBLIOGRAPHY
duality functor, 303 generic splitting tower, 102

graded Milnor ring, 406
Euler class, 246 graph of a morphism, 296
Excess Formula, 255 Grothendieck-Witt group
excess vector bundle, 255 presentation of, 16
exchange isomorphism, 296 group
exponent, 396 of p-dimensional cycles, 265
external product, 234 of projective similitudes, 314
field of similitudes, 314
ū-invariant of, 161 Gysin homomorphism, 254, 269
ūr (F )-invariant of, 162 Harrison subbasis, 392
p-generic, 375 Hasse number, 173
u-invariant of, 161 linked fields, 174
u0 -invariant of, 166 Hasse Principle
p-special, 414 strong, 173
admissible extension of, 125 weak, 173
euclidean, 123, 390 Hasse-Witt invariant, 24
finite stable range, 181 Hauptsatz, 29, 98
formally real, 70, 390 Herbrand index, 416
function field of a quadric, 92 homomorphism:first residue, 78
generic, 375 homomorphism:second residue, 78
Hasse number of, 173 Homotopy Invariance, 246, 269
isotropy, 375 hyperbolic λ-bilinear form, 4
Laurent series field, 124 hyperbolic bilinear form, 4
leading field of a quadratic form, 103 hyperbolic form, 34
linked, 173 hyperbolic pair, 4, 34
ordered, 390 hyperbolic plane, 34
existence and uniqueness of a real bilinear, 4
closure, 391 hyperplane reflection, 34
extension of, 390
pythagorean, 123 idempotent completion of an additive cat-
pythagorean closure of, 125 egory, 302
quadratic closure of, 125 index of an algebra, 396
quadratically closed, 121 intersecting properly subvarieties, 271
real closed, 390 isometry
real closed with respect to an order- bilinear form, 3
ing, 390 quadratic form, 34
reduced stability index of, 135 Izhboldin dimension, 331
relative u-invariant of, 167
stable range of, 181 level of a field, 26
field extension Local-Global Principle, 128
excellent, 112 localization exact sequence, 243
Frobenius Reciprocity, 79 Localization Sequence, 268
locally hyperbolic, 161
Galois module, 401 locally isotropic, 173
INDEX 443
matrices quadratic, 48
congruent, 4 r-linked, 101
matrix chain p-equivalent, 60
alternating, 4 characterization of, 97, 108
Milnor Conjecture, 64 general, 48
Milnor’s Theorem, 407 linkage of, 101
module linked, 101
m-periodic, 404 neighbor, 99
monomial pure subform, 62
degree simply p-equivalent, 60
in lexicographic order, 71 Pfister neighbor, 99
motive, 300, 303 associated general Pfister form, 99
multiplier, 313 place, 416
center of, 417
negative with respect to an ordering, 390 composition, 417
Nilpotence Theorem, 306, 381 geometric, 417
norm, 394 induced by, 417
norm homomorphism, 408 surjective, 416
norm residue homomorphism, 412 polar identity, 33
normal bundle, 420 polynomial
Normality Theorem, 136
σ-twisted, 399
order homomorphism, 266 marked, 230
order in a ring, 115 leading coefficient of, 71
ordering, 390 leading term of, 71
orientation, 313, 317 positive with respect to an ordering , 390
orthogonal basis, 6 preordering, 390
orthogonal complement, 5 prime cycle, 265
orthogonal subspace, 5 Projection Formula, 262
orthogonal sum projective bundle, 418
external, 6, 38 Projective Bundle Theorem, 249, 269
internal, 5, 36 proper inverse image, 270
orthogonal vectors, 5 Property An , 149
going down, 152
Pfister form going up a quadratic extension, 150
bilinear, 17 Property Hn , 173
r-linked, 30 pull-back homomorphism, 225, 242, 258,
chain p-equivalent, 26 268
divides, 28 pure quaternion, 397
general, 28 push-forward homomorphism, 222, 242,
linkage, 30 267
linkage number, 30 pythagoras number, 338
linked, 30
pure subform of, 27 quadratic form, 33
simply p-equivalent, 26 nth cohomological invariant of, 63
divisor of, 100 qth kernel form, 102
444 BIBLIOGRAPHY
absolute higher Witt index, 102 Quadratic Similarity Norm Principle, 82

anisotopic, 35 Quadratic Similarity Theorem, 83
anisotropic part, 43 Quadratic Substitution Principle, 68
annihilators of, 50 Quadratic Value Norm Principle, 74
associated, 34 Quadratic Value Theorem, 72
Clifford invariant, 59 quadratic Witt group, 46
cohomological invariant of, 63 Principle ideals generated by Pfister
complementary form, 36 forms, 183
defined over a field, 110 quadratically closed field, 9
degree of, 103 quadric
determinant, 58 oriented, 313, 317
diagonalizable, 39 projective, 91
dimension, 33 split, 311
discriminant, 57 quadric form
excellent, 112 subform, 35
group of similarity factors, 47 quasi-Pfister form, 52
height of, 102
radical
irreducible, 92
bilinear, 5
isotropic, 35
ramification index, 225
kernel form, 102
rationally equivalent cycles, 267
leading form of, 103
reduced Chow group, 317
non-degenerate, 37
reduced norm map, 397
polar form, 33
reduced stability index, 135
polar form of, 75
reduced trace map, 397
regular, 36
relative u-invariant, 167
regular part, 36 relative dimension, 225
relative higher Witt indices, 103 relative tangent bundle, 420
round, 47 Representation Theorem, 69
splits off, 37 represented values, 7, 47
splitting pattern of, 102 residue homomorphism, 220, 407
tensor product, 45 restriction graded ring homomorphism, 406
totally isotropic subspace, 35 restriction homomorphism, 226, 403, 411
totally singular, 36 restriction map
universal, 47 bilinear forms, 14
quadratic forms quadratic forms, 46
classification of binary quadratic forms, ring
55 artinian, 415
non-degenerate excellent, 219
characterizations, 37, 54 Rost correspondence, 347
totally singular
characterization of, 51, 52 scheme, 219, 265, 281
quadratic norm form of a quadratic étale rank of, 381
F algebra, 48 relatively cellular, 304
quadratic radical, 36 tractable, 381
INDEX 445
Segre closed embedding, 396 Tate motive, 300, 303

Segre homomorphism, 273 tautological
Separation Theorem, 107 line bundle, 422
sequence tensor category, 301
regular, 419 torsion-free index, 191
regular of codimension d, 419 total Chern class, 251
Severi-Brauer variety, 396 total Euler class, 272
generalized, 396 total Segre class, 273
shell triangle, 324, 326 total Segre operation, 274
signature, 123 total Stiefel-Whitney class, 22
total, 134 total Stiefel-Whitney map, 22
cokernel of, 135 ith, 23
signature map, 123, 128 totally indefinite, 173
total, 134 totally negative, 390
signed discriminant, 58 totally positive, 390
similar trace, 394
bilinear forms, 4 tractable class, 381
quadratic, 35 transfer
similitude, 313 of bilinear forms, 79
simply chain equivalence of quadratic forms, 79
simply transfer maps
bilinear forms, 14 of Witt rings and groups, 80
space of orderings, 390 transpose of a correspondence, 296
specialization, 220
specialization homomorphism, 408 variety, 219, 220, 265
specialization homomorphisms, 238 vector
split class, 396 anisotropic, 9, 35
split idempotent, 302 isotropic, 9, 35
split object, 303 polynomial
splitting field extension, 396 degree of, 71
Splitting Principle, 250 leading term of, 71
Springer’s Theorem, 73, 317 leading vector of, 71
stable range see field primitive, 75
stable range of 181 Vector bundle
Steenrod operations tautological, 364
of cohomological type, 281, 290 vector bundle, 418
of homological type, 281
Strong Splitting Principle, 280 weight, 368
Subform Theorem, 92 Weil’s Reciprocity Law, 224
Sylvester’s Law of Inertia, 123 Whitney formula, 23
symbol, 406 Whitney Sum Formula, 253
symmetrization, 350 Witt classes, 12
symplectic basis, 6 Witt equivalent, 43
Witt group
tangent bundle, 420 presentation of, 16
446 BIBLIOGRAPHY
quadratic, see quadratic Witt group

Witt index, 43
of a quadric, 91
Witt relations, 15
Witt ring, 12
graded, 21
annihilator ideals in, 30
annihilators in, 182
exact sequence for quadratic field ex-
tensions, 139, 140, 145, 147
exact sequence on the affine line, 85
exact sequence on the projective line,
87
fundamental ideal, 17
isomorphism of, 130
Principle ideals generated by Pfister
forms, 183
reduced, 137
Spectrum of, 128
structure of in the formally real case,
129
structure of over non formally real fields,
122
Witt theorems
Bilinear Witt Cancellation Theorem,
11
characteristic not two, 8
characteristic two counterexample,
8
Bilinear Witt Chain Equivalence The-
orem, 14
Bilinear Witt Decomposition Theorem,
11
Quadratic Cancellation Theorem, 43
Quadratic Witt Decomposition The-
orem, 43
Witt theorems:(Quadratic) Witt Exten-
sion Theorem, 41
Witt-Grothendieck ring, 12
Wu Formula, 290
zero section, 418
NOTATION 447
Notation
(L/F, b), quaternion algebra, 397 F (Xϕ ) function field of Xϕ if ϕ irre-

(a), 402 ducible, 92
(a)m , 402 F (p) quotient field F [T ]/(p), p ∈ F [T ]
A?B, operation on quadratic étale irreducible, 71
algebras, 395 F (t) rational function field over F in
p
A (X, Kq ), K-homology groups of one variable, 9
X, 260 F (x), residue field of x, 220
Aop , opposite algebra, 396 F [T ] polynomial ring over F in n
AL , A ⊗F L, 394 variables t1 , . . . , tn , 67
Ap (X, Kq ), K-homology group of F [t] polynomial ring in variable t over
X, 242 F, 8
BX , blow up BX along ∆(X) × F [t1 , . . . , tn ] polynomial ring over F
{0}, 284 in n variables, 67
C(X), Rost complex for X, 221 F× unit group of F , 4
×2
C(ϕ) Clifford algebra of ϕ, 52 F group of nonzero squares in F ,
¡ ¢ 4
C X(F ), 2 Z ¡ ring of continuous
m
¢ func-
FP real closure of F at the ordering
tions C X(F ) → 2 Z , 134
m
0
P , 128, 391
C ⊕C , direct sum of cones, 418 Fa quadratic étale F -algebra deter-
C + (ϕ) ¡ + ¢simple algebra
central with C(ϕ) ' mined by a, 48
2
M2 C (ϕ) , ϕ ∈ Iq (F ), 59 Fpy pythagorean closure of F , 125
◦
C , cone with zero section deleted, Fsep , separable closure of F , 401
421 Fh(ϕ)−1 leading field of ϕ, 103
C0 (ϕ) even Clifford algebra of ϕ, 53 G(b) group of similarity factors of b,
CQ , conic curve, 194 7
Cf , normal cone of f , 418 G(ϕ) group of similarity factors of ϕ,
Cp (X), 221
¡ 47 ¢
Cp,n (X), 221 GC X(F ), Z¡ graded
¢ ring associated
D(t) D ⊗F F (t), D an F -division al- to C X(F ), Z , 184
gebra, 115 GP (F ) set of general quadratic Pfis-
D(b) set of non-zero values of b, 7 ter forms, 48
D(ϕ) nonzero values of ϕ, 47 GPn (F ) set of general quadratic n-
D[t] D ⊗F F [t], D an F -division al- fold Pfister forms, 48
¡ gebra,
¢ 115 GW∗ (F ) graded Witt group associ-
D ∞h1i group of nonzero sums of atied to the fundamental ideal, 21
squares in F , 123 H(a) orderings P with a 6∈ P , 392
Df , deformation scheme, 423 H(a1 , . . . , an ) = HF (a1 , . . . , an ) ∩H(ai ),
F ((t)) quotient field of the ring of 392
power series F [[t]], 124 H n (F ) H n,n (F, Z/2Z), 63
F (T ) rational function field over F n
H (F ), H n,n (F, Z/2Z), 412
in n variables t1 , . . . , tn , 67 H n (F, M ), H n (ΓF , M ), 401
448 BIBLIOGRAPHY
H n (Γ, M ), nth cohomology group Pn (F ) set of quadratic n-fold Pfister

of Γ with coefficients in M , 401 forms, 48
H n,i (F, Z/mZ), 410 R[t, σ], 399
I(F ) fundamental ideal (of even di- R× unit group of a ring R, 4
mensional forms ¡ in W ¢n (F )), 16 TX , tangent cone of X, 419
I n (F ) nth power I(F ) of the fun- Tg , relative tangent bundle of g,
damental ideal, 17 420
n m UX , 282
I J , 424
Iq (F ) quadratic Witt group, 46 V (E), 210
Iq (K/F ) kernel of rK/F : Iq (F ) → V (T ) F (T ) ⊗F V , 67
Iq (K), K/F , 98 V (t) F (t) ⊗F V , 67
Iqn (F ) I n−1 (F ) · Iq (F ), 48 V [T ] F [T ] ⊗F V , 67
It (F ) Wt (F ) ∩ I(F ), 130 V [t] F [t] ⊗F V , 67
Itn (F ) Wt (F ) ∩ I n (F ), 148 VK base extension of V to K, 5
Ib (F ) I(F ) ∩ annW (F ) (b), 154 VQ , space of pure quaternions, 194
n
Ired (F ) image of I n (F ) in Wred (F , Vϕ underlying vector space of ϕ, 33
138 W (F ) Witt ring of F , 12
J(ϕ), J-invariant, 372 W (K/F ) kernel of rK/F : W (F ) →
K(ϕ) K(ϕK ) if ϕ is irreducible, K/F , W (K), K/F , 98
92 W⊥ orthogonal complement of W , 5
K * L, place, 416 Wt (F ) additive torsion subgroup of
K∗ (F ), graded Milnor ring of F , W (F ), 126
406 Wred (F reduced Witt ring, 137
K0 (X), Grothendieck group of the Xϕ projective quadric associated to
category of vector bundles over ϕ, 91
X, 418 Xϕ (L) set if L=points of Xϕ , L/F ,
Kn (F ) Milnor nth Kn -group, 21 91
L(D), line bundle associated with Xq Xϕq , 103
divisor D, 271 X(p) , set of point of X of dimen-
Lc , canonical line bundle over P(C), sion p, 220
421 Y /G, factor scheme, 428
G
Ln , 199 Y , 427
Lt , tautological line bundle over Yx , fiber scheme over x, 225
P(C), 421 Z(a), 266
M (−1), 404 [A] · [B], product in Br(F ) and
M (X), Chow motive of X, 300 Ét2 (F ), 395
M (X)(i), Chow motive (X, i), 300 [Z], cycle class of Z, 242
Γ [Z], cycle of Z, 221
M , subgroup of Γ-invariant ele-
ments of M , 401 [a, b] quadratic form on F 2 given by
N (ρ) dim ρ − 2deg ρ , 108 ϕ(x, y) = ax2 + xy + by 2 , 34
Nf , normal bundle of f , 420 [x],
¡ ¢◦ cycle of a point,0 265
P (F ) set of quadratic Pfister forms, b ⊗ hhd]] the form b ⊗ hhd]], 62
48 Br(F ) Brauer group of F , 59
P ∨, dual sheaf, 418 Br(F ), Brauer group of F , 395
NOTATION 449
Br(L/F ), 396 annm (ρ) elements in I m (F ) annihi-

Br2 (F ) 2-torsion subgroup of Brauer lated by ρ, 182
group of F , 59 annm (b) elements in I m (F ) annihi-
Brm (F ), 396 lated by b, 181
CH(Z; Λ), Chow group of classes annW (F ) (ϕ) annihilator of ϕ in W (F ),
of cycles with values in Λ, 299 50
CH∗ (X), graded Chow group, 271 annW (F ) (b) annihilator of b in W (F ),
p 30
CH (X), Chow group of codimen-
sion p classes of cycles on X, 271 ē0 map induced by the dimension map
CHp (X), Chow group of dimen- on W (F )/I(F ), 17
sion p cycles on X, 265 ē1 map induced by the signed deter-
char F characteristic of F , 3 minant, 18
Ét2 (F ), group of isomorphism classes ū(F ) ū-invariant of F , 161
of quadratic étale F -algebras, 395 ūr (F ) ūr -invariant of F , 162
ΓF , absolute Galois group of F , β
◦ α, ® bilinear pairing, 296
401 D(ϕ) the group generated by D(ϕ),
Γf , graph of a morphism f , 296 72
Grr (V ), Grassmannian variety of OXϕ (s), 313
r-dimensional subspaces in V ., 396 S • (V ) symmetric algebra of V , 91
IndK/F (M ), 404 cdp (F ), cohomological p-dimension
Λ(i), Tate motive with values in of F , 415
Λ, 300 χ(ΓF ), character group of ΓF , 401
NA quadratic norm form of the qua- χm (ΓF ), 401
dratic F -algebra A, 55 clif(ϕ) Clifford invariant of ϕ, 59
NA , norm map, 394 deg ϕ degree of ϕ, 103
Nrd, reduced norm, 194, 397 deg f leading coefficient of f ∈ F [T ]
P(C), projective cone of C, 418 in lexicographical order, 71
SB(A), Severi-Brauer variety, 396 deg v leading coefficient of v ∈ V [T ]
TrA trace form of the quadratic F - in lexicographical order, 71
algebra A, 55 deg, degree homomorphism, 267
TrA , trace map, 394 deg(A), degree of A, 396
Trd, reduced trace, 194, 397 δ, connecting homomorphism, 243
Z(X; P1 ), 267, 283 det b determinant of b, 4
Zp (X), group of p-dimensional cy- dim x, dimension of of x, 220
cles on X, 265 disc(ϕ) discriminant of ϕ, 57
µ ¶
α : X Ã Y, correspondence be- a, b
tween X and Y , 299 quaternion algebra in char-
F
α ∪ β, cup-product, 402
µ ¶ acteristic not two, 49
α × β, external product, 234 a, b
∗ , quaternion algebra, 397
α , 298 F
2 · ¸
αG , 282 a, b
αt , transpose of α, 296 quaternion algebra in charac-
F
α∗ , 297 teristic two, 49
450 BIBLIOGRAPHY
· ¸
a, b CR∗ (F ), category of correspondences
, quaternion algebra, 398
F over F , 301
div(f ), divisor of f , 221 CR∗ (F, Λ), category of correspon-
ind(A), index of A, 396 dences with coefficients in Λ, 300
κ(x), residue field of x, 220 Corri (X, Y ; Λ), group of correspon-
haib one dimensional bilinear form b(v, w) = dences between X and Y , 299
avw, 7 C(F, Λ), 300
haiq one dimensional quadratic form GO(ϕ), group of similitudes of ϕ,
ϕ(v) = av 2 , 34 314
ha1 , . . . , an ib orthogonal sum of bilin- Gras(ϕ), 363
ear forms hai i (also written ha1 i ⊥ PGO(ϕ), group of projective simil-
· · · ⊥ han i), 7 itudes of ϕ, 314
ha1 , . . . , an iq Sg(C),
orthogonal sum of ha1 i, . . . , han i, total Segre class, 273
39 Sm, category of smooth complete
schemes, 295
hhaiib h1, −aib (also written hhaii),
16 SqX , Steenrod operation of homo-
logical type, 285
hha1 , . . . , an iib bilinear n-fold Pfister
SqX , Steenrod operation of coho-
form hha1 ii ⊗ · · · ⊗ hhan ii (also
mological type, 290
write hha1 , a2 , . . . , an ii, 17
SqkX , Steenrod operation of coho-
hha1 , . . . , an ]] quadratic n-fold Pfister
mological type, 290
form hha1 , . . . , an−1 iib ⊗ hhan ]], 48
SqX k , Steenrod operation of homo-
hha]] norm form of the quadratic étale logical type, 285
F -algebra Fa , 48
O(V, ϕ), orthogonal group of ϕ,
hha1 , . . . , an iiq (quadratic) quasi-Pfister 363
form hha1 , . . . , an iib ⊗ h1iq , 52 +
O (V, ϕ), special orthogonal group
1, trivial line bundle, 418 of ϕ, 363
µa , 197 cdim(X), canonical dimension of
µm , group of mth roots of unity, X, 375
402 cdim(C), canonical dimension of
ν(F ) torsion-free index of F , 191 a class C, 375
SBr (A), generalized Severi-Brauer cdimp (X), canonical p-dimension
variety, 396 of X, 375
CM* (F, Λ), category of Chow mo- cdimp (C), canonical p-dimension
tives with coefficients in Λ, 303 of a class C, 375
CM(F ) category of graded Chow ordR , order homomorphism, 266
motives over F , 303 ordx , order homomorphism asso-
CM(F, Λ), category of graded Chow ciated with a point x, 266
motives with coefficients in Λ, 303 sgC , Segre homomorphism, 273
n n
CR(F ), category of graded cor- I (F ) I (F )/I n+1 (F ), 143
m
respondences over F , 301 annm (ρ) elements in I (F ) annihi-
CR(F, Λ), category of graded cor- lated by ρ, 182
m
respondences with coefficients in annm (b) elements in I (F ) annihi-
Λ, 300 lated by b, 181
NOTATION 451
en : Iqn (F )/Iqn+1 (F ) → H n (F ) nth ϕ1 ⊥ ϕ2 (external) orthogonal sum

graded cohomological invariant, 64 of ϕ1 and ϕ2 , 38
∂ : W (K) → W (K̄) first residue ho- ϕK base extension of ϕ to K, 37
momorphism, 77 ϕq qth kernel form of ϕ, 102
∂xx0 , residue homomorphism, 220 ϕan anisotroic part of ϕ, 43
∂R , residue homomorphism, 220 ϕb quadratic form associated to b, 34
∂ZU , boundary map, 232 ϕh(ϕ)−1 leading form of ϕ, 103
∂π : W (K) → W (K̄) second residue b
I(F ) kernel of the dimension map on
homomorphism, 77 the Witt-Grothendieck ring, 16
¡ ¢
∂v , residue homomorphism, 405, Ibn (F ) nth power I(F b ) n of I(F
b ), 17
407 c (F )
W Witt-Grothendieck ring of F ,
∂x , residue homomorphism, 405,
12
407 e
D(∞h1i) sums of squares in F , 390
ψ⊥ complementary form of a subform e
ψ, 36 D(b) D(b) ∪ {0}, 7
e
D(ϕ) D(ϕ) ∪ {0}, 47
rad ϕ quadratic radical of ϕ, 36 ¡ ¢ ¡ ¢
rad b radical of b, 5 De ∞h1i D ∞h1i ∪ {0}, 123
sgn b : X(F ) → Z total signature e
G(ϕ) G(ϕ) ∪ {0}, 52
map of b,¡ 134 ¢ u
e(F ) Hasse number of F , 173
sgn : W (F ) → C X(F ), Z total sig- ℘ Artin-Schreier map, 35
nature map, 134 ℘, Artin-Schreier map, 395
sgn : W (F ) → Z signature map, 123 {a1 , . . . , an } n-symbol in Milnor Kn (F )
sgn b signature of b, 123 and kn (F ), 21
sgnP : W (F ) → Z signature map at {a}, endomorphism of C∗ (X), 221
the ordering P , 128 {a1 , a2 , . . . , an }, endomorphism of
σf , deformation homomorphism, C∗ (X), 221
238, 243 {a1 , a2 , . . . , an }, symbol in Kn , 406
st(F ) stable range of F , 181 {a1 , a2 , . . . , an }F , symbol in Kn (F ),
str (F ) reduced stability index of F , 406
a
135 aM , Ker(M − → M ), 416
supp b support of b, 135 bn , Bockstein map, 414
τv (hyperplane) reflection about v, c(E), total Chern class, 251
34 ci (E), Chern classes, 251, 275
Hom(M, N ), 304 cK/F , corestriction homomorphism,
ϕ ≺Â ψ ϕF (ψ) is isotropic, 95 403
ϕ ≺Â ψ ϕ Â ψ and ψ Â ϕ, 95 cL/F , corestriction homomorphism,
ϕ∼ψ ϕ and ψ are Witt equivalent, 411
43 cL/F , norm homomorphism, 408
ϕ'ψ ϕ and ψ are isometric, 34 dX , differential of C∗ (X), 221
ϕ, φ, ρ, etc quadratic forms, 33 e(E), Euler class of E, 246
0 •
ϕ pure subform of the quadratic Pfis- e(L) , total Euler class, 272
ter form ϕ, 62 e0 dimension map mod 2 on the Witt
ϕ |W restriction of ϕ to W , 35 ring, 17
452 BIBLIOGRAPHY
e1 : Iq (F ) → Ét2 (F ) discriminant rL/F , restriction homomorphism,

map, 58 396, 406, 411
e1 map induced on I(F ) by the signed s(E), total Segre operation, 274
determinant, 17 s(F ) level of F , 26
e2 map I 2 (F ) → k2 (F ), 22 s∗ : Iq (K) → Iq (F ) transfer map
e2 : Iq2 (F ) → Br2 (F ) Clifford invari- induced by the linear functional
ant map, 59 s : K → F , K/F finite, 80
en : Iqn (F ) → H n (F ) nth cohomolog- s∗ : W (K) → W (F ) transfer map
ical invariant, 63 induced by the linear functional
¡ ¢
en hha1 , . . . , an ]] cohomological invari- s : K → F , K/F finite, 80
ant of hha1 , . . . , an ]], 63 s∗ : W c (K) → W c (F ) transfer map
ey (g), 225 induced by the linear functional
f∗ leading coefficient of f ∈ F [T ], 71 s : K → F , K/F finite, 80
f ∗, pull-back homomorphism, 258 s∗ (b) transfer of b induced by the lin-
F
f , Gysin homomorphism, 254, ear functional s : K → F , K/F
269 finite, 79
f∗ map taking a k∗ (F ) → GW∗ (F ), sπ , specialization homomorphism,
21 408
f∗ , push-forward homomorphism, u0 (F ) u0 -invariant of F , 166
222, 242, 267 u(F ) u-invariant of F , 161
g∗, pull-back homomorphism, 225, u(L/F ) relative u-invariant of L/F ,
242, 268 167
h(a, M ), Herbrand index, 416 uX , 284
n
hF , norm residue homomorphism v∗ leading coefficient of v ∈ V [T ], 71
of degree 2, 412 vX , 283
¢
hn,m c (F ) → k∗ (F ) [[t]]×
F , norm residue homomorphism, w:W total Stiefel-
412 Whitney map, 22
hnF norm residue homomorphism of wi ith total Stiefel-Whitney map, 23
degree n, 64 x <P 0 x 6∈ P , 390
i0 (Xϕ ) Witt index of Xϕ , 91 x >P 0 x ∈ P , 390
kn (F ) Milnor Kn (F )/2Kn (F ), 21 Mn (F ) n × n matrix ring over F , 3
l(A), length of A, 415 S 2 (V ) symmetric square of V , 4
lA (M ), length of M , 415 H hyperbolic plane, 34
la , linear function on VQ , 195 H(V ) hyperbolic form on V , 34
nϕ orthogonal sum of n copies of ϕ, Hλ hyperbolic λ-bilinear plane, 4
39 Hλ (V ) hyperbolic λ-bilinear form on
nb orthogonal sum of n copies of b, V, 4
6 Q, R, C field of rational numbers, real
rK/F restriction map induced by K/F number, complex numbers, respec-
on Witt(-Grothendieck) ring, 13 tively, 98
rK/F , restriction homomorphism, Z the ring of integers, 13
403 X(F ) set of orderings of F , 390
rK/F : Iq (F ) → Iq (K) restriction map b ∈ W (F ) interpreted as the Witt
of quadratic Witt groups, 46 class of b in W (F ), 12
NOTATION 453
b⊗ϕ tensor product of b and ϕ, 45

b0 pure subform of a bilinear Pfister
form b, 27
b, c, d, e, etc bilinear forms, 3
b|W restriction of b to W , 4
b1 ' b2 isometry of b1 and b2 , 3
b1 ⊗ b2 tensor product of b1 and b2 ,
12
b1 ⊥ b2 external orthogonal sum of
b1 and b2 , 6
bK base extension of b to K, 5
bϕ polar form of ϕ, 33
ban anisotropic part of b, 11
h(ϕ) height of ϕ, 102
i(b) Witt index of b, 11
i0 (ϕ) Witt index of ϕ, 43
iq (X) iq (ϕ), 103
iq (ϕ) qth relative higher Witt index
of ϕ, 103
jq (X) jq (ϕ), 103
jq (ϕ) qth higher absolute Witt index
¡ of¢ϕ, 102
nil¡ W (F )¢ nilradical of W (F ), 122
zd W (F ) zero divisors of W (F ), 122
Jn (F ) set of even non-degenerate qua-
dratic forms of degree at least n,
104

3892 PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

3892 PDF

Uploaded by

Copyright:

Available Formats

The Algebraic and Geometric Theory of

Richard Elman, Nikita Karpenko,

Part . Classical theory of symmetric bilinear forms and quadratic forms 1

28. Excellent quadratic forms 110

Part . Algebraic cycles 219

60. Properties of the Steenrod operations 287

Part . Quadratic forms and algebraic cycles 311

93. Criterion of isomorphism 385

Decomposition. The Witt group Iq (F ) of non-degenerate even dimensional quadratic

Classical theory of symmetric bilinear forms

The study of (n × n)-matrices over a field F leads to various classification problems.

form based on V . If b is symmetric (respectively, alternating) then the associated matrix

1.A. Structure theorems for bilinear forms. Let b be a symmetric or alternating

Proposition 1.7. Let b be a symmetric or alternating bilinear form on V . Let W be

Example 1.15. Let K = F [t]/(t2 − a) with a ∈ F , where F [t] is the polynomial

if a ∈ F ×2 and ha, ai ' hb, bi if and only if aF ×2 ' bF ×2 .

Corollary 1.10, we have c2 = 0 if char F 6= 2 and c2 = mH1 for some integer m if

1.C. Metabolic bilinear forms. Let b be a symmetric bilinear form on V . A

(3) A 2-dimensional metabolic space is nothing but a non-degenerate isotropic plane.

Theorem 1.28. (Bilinear Witt Decomposition Theorem) Let b be a non-degenerate

2. The Witt and Witt-Grothendieck rings of symmetric bilinear forms

Exercise 2.3. (Cf. Scharlau [125], p.22.) Let b be a non-degenerate symmetric

The Witt Cancellation Theorem 1.29 allows us to conclude the following.

Proposition 2.7. A homomorphism of fields F → K induces ring homomorphisms

Two non-degenerate diagonal symmetric bilinear forms a = ha1 , a2 , . . . , an i and b =

4. Structure of the Witt ring

4.A. The presentation of W c (F ) and W (F ). We turn to determining presentations

If x = 0 or y = 0, the statement is obvious, so we may assume that x, y ∈ F × . It follows

Proof. It suffices to prove that the homomorphism W 0 (F ) → W c (F ) taking [a] to

By Corollary 1.25, we have a map

The map e1 is a homomorphism

(4) hha, bii + hha, −bii = hha, −1ii in W (F ).

derived from Lemma 4.15, followed by using equation (4.19) yields

0 −−−→ I 2 (F ) −−−→ I(F ) −−−→ I(F )/I 2 (F ) −−−→ 0

where f1 is the isomorphism inverse of ē1 in Proposition 4.13.

Example 5.5. If b is a metabolic plane then b = hai + h−ai in W c (F ) for some

w(α) = (1 + {a1 , . . . , an , −1, . . . , −1 }tm )−1 .

Let w0 (α) = 1 and

Remark 5.8. Let K/F be a field extension and α ∈ W c (F ). Then

This gives another proof for Propositions 4.13 and 5.3.

The following fundamental theorem was proved by Orlov-Vishik-Voevodsky [110] in

6. Bilinear Pfister forms

The isometry classes of tensor products of non-degenerate binary symmetric bilinear

Proof. We induct on n. If n = 1 then −c = yb1 for some −y ∈ D(b) and this

A non-degenerate symmetric bilinear form b is called a general bilinear n-fold Pfister

Proof. Let a = hha1 , . . . , an ii and b = hhb1 , . . . , bn ii be isometric over F . Clearly

We need the following theorem of Arason and Pfister (cf. [11]):

6.B. Linkage of bilinear Pfister forms. We derive some further properties of

If a general bilinear r-fold Pfister form c is isometric to a common subform of two

Proof. We prove this by induction on dim c. By the Witt Decomposition Theorem

If ϕ, ψ ∈ Quad(V), we say ϕ is similar to ψ if there exists an a ∈ F × such that ϕ ' aψ.

Example 7.9. If F is algebraically closed then any homogeneous polynomial in more

Let ψ be a subform of a quadratic form ϕ. The restriction of ϕ on (Vψ )⊥ is denoted

7.A. Non-degenerate quadratic forms. To define non-degeneracy, we use the fol-

Let ϕi be a quadratic form`on Vi for i = 1, 2. Then their external orthogonal sum is

In particular, if n is an integer, we do not interpret nϕ with n viewed in the field. For

(2) If dim ϕ = 2n + 1 then

W0 ⊂ (F u)⊥ . Applying the induction hypothesis to β = α|W0 : ϕ|W0 → ϕ|α(W0 ) , there

Proposition 8.8. Let ρ be a non-degenerate quadratic form of even dimension over

8.B. Totally isotropic subspaces. Given a regular quadratic form ϕ on V , we

(2) If either ϕ or b is hyperbolic then b ⊗ ϕ is hyperbolic.

As the orthogonal sum of even dimensional non-degenerate quadratic forms over F is

ϕ'ψ if and only if ϕ = ψ in Iq (F ) and dim ϕ = dim ψ.